The Best Ice Cube Tray

The best ice cube tray around for cocktails and other chilled beverages is Tovolo's King Ice Cube Tray. Literally every bartender and mixologist we've spoken to has recommended it—by name. Its large cubes are great for most drinks and it's easy to use and clean. It ranges from about $7.50 to $12, depending on which color you choose; blue and red happen to be the cheapest on Amazon.

Last Updated: October 9, 2013
Added long term test results, and solutions for removing smells and hard water residue from the cube trays.
Expand Previous Updates
May 29, 2013: Performed a blind taste test after receiving some feedback that maybe big ice cubes don't matter that much in terms of taste.  They do, but it's not a huge deal.

Ice may seem like the most straightforward recipe in your kitchen, but frozen water deserves at least a little more thought than the average person gives it. It’s time to get out of the mindset of those cheap, plastic trays that come with your freezer. Ice is an extremely important ingredient—and often overlooked—when it comes to making quality beverages. For example, The Aviary, 3-Michelin Star chef Grant Achatz’s cocktail bar that specializes in advanced molecular gastronomy, treats every drink as a plated dish, produces literally dozens of different kinds of ice every night, and chips away at 300 pound blocks to get the exact size and shape they need for each specific use, like bartenders used to before Prohibition. Obviously, this isn’t going to happen at home. You need a simple vessel to pour water into, and get ice out of.

Cheat sheet: Different ice for shaking vs. stirring vs. presentation

OK, so what type of ice should you use, and when? I’ll keep this simple.

For cocktail shaking or drinking water or other things you want cooled fast without worry of dilution. Ice from the Perfect Cube Silicone Ice Cube Tray, as well as most regular trays or the half-moon ice from your freezer ice-maker are fine here (all make around 1-inch cubes). According to tests by Dr. Delicious (a.k.a.  Dave Arnold), all ice cubes are pretty much equal when it comes to chilling down liquids in a cocktail shaker. Larger cubes are harder to shake and crushed ice introduces too much surface water.

For cocktails on the rocks.  Here’s where I break out the big cubes (Tovolo’s King Ice Cubes) and spheres. If you want a cold drink, with some water dilution, big ice is the way to go, as the cooling/diluting process will happen more slowly, giving you more time in the optimal temperature and dilution zone. For a cocktail or whiskey drink you could also use whiskey stones, but they won’t get your drink as cool and they won’t dilute your drink, which could be good or bad depending on what your drink preferences are. Also remember that the presentation of a food or drink can significantly impact the pleasure we derive from it. That’s why Scotch and wine tasters always begin by looking at their spirits, not tasting them. A big, clear hunk of ice is undeniably beautiful.

What makes a great ice cube tray for cocktails?

Any ice cube tray can do that, though. What separates a great one from the no-name plastic trays is cube shape, size, and ease of ejection. The first two factors have a big effect on how the ice melts in your glass, and how quickly it dilutes your drink. Brian Van Flendern was the head mixologist at Thomas Keller’s Per Se, a 3-Michelin Star restaurant which is widely regarded among the best in the world, and now runs Creative Cocktail Consultants. He recommends the Tovolo. “For the home user I’d recommend [it]…a larger cube makes for better cocktails.”

The science and practical nature of whether or not this makes a difference is marginally important, though. Big ice cubes also look great in drinks, and that’s how the pros serve cocktails when they want something that will dilute a drink slower, they told us. Flendern continued, “You have a slower rate of dilution”. We spent about ten hours researching, interviewing almost a dozen different bartenders, cocktail experts, and scientists to find the best choice. Then we did our own casual tests and the data backed up their choice.

Our pick

Aside from the fact that every professional drink maker we spoke to loves this silicone ice-mold, there are two main reasons why King Ice Cube Tray is our top pick. First, the cubes themselves. Larger cubes look cool, sure, but more importantly, they melt more slowly than smaller ones, meaning your drink won’t get diluted as quickly. “They dissolve more slowly due to their smaller surface area to volume ratio than the same mass of smaller cubes,” says John Seddon, professor of chemical physics at Imperial College London.

For reference, Lustroware makes a Covered Ice Cube Tray, available at The Container Store. Each cube is about 10 grams, and has a surface area of 7 square inches—about half to a third the size your average ice cube. On the other hand, King Ice Cube Tray creates ice with a surface area of roughly 24 square inches and a mass of around 100 grams. This means that the same mass of the smaller cubes has about three times as much surface area as the 2-inch cubes.

The second reason is the silicone rubber material from which the King Ice Cube Tray is made. It’s easier to remove ice from silicone than it is to pry it from plastic or metal. “The silicon-oxygen backbone of the polymer has strong chemical bonds and is also very flexible; the methyl side groups prevent crystallization, keeping it malleable, and also confer non-stick properties due to their hydrophobic and non-reactive character,” says Seddon. Instead of having to crack back and forth to loosen the ice, you simply push it out. You don’t have to worry about every other cube coming out when you only want one. What’s nice about the Tovolo King is that the silicone is rigid enough to support the shape without being floppy. (That’s one of the problems with Casabella’s Silicone Big Ice Trays.) Additionally, the Tovolo King is dishwasher safe, and BPA-free, which is always a plus.

“Speakeasy Cocktails: Learn from the Modern Mixologists” is an essential iPhone/iPad app for anyone interested in mixing up cocktails. In the app, contributors Jim Meehan, who has been named “American Bartender of the Year” by Tales of the Cocktail Spirit, and whose bar, PDT (Please Don’t Tell), has earned the distinction of “World’s Best Cocktail Bar” from the same organization, and Joseph Schwartz, co-founder of Little Branch and one of Food & Wine’s top five bartenders in America, share their knowledge on not only techniques and recipes, but tools, too. The app says the shape King Ice Cube Tray puts out is great for what’s called ‘shaking ice.’ “Its bulk allows bartenders to shake vigorously without risking excess dilution.” Additionally, the authors prefer it for “strong cocktails like the Old Fashioned, where minimal melting is preferred.”

“Tovolo was the first to release these types of molds, and I think they’re great,” says Meehan, although he now uses comparable models from (Other bartenders have told us the trays from Cocktail Kingdom aren’t as sturdy as Tovolo’s, and they otherwise didn’t come up in our research.)

We spoke to a number of bartenders, and they all gave the Tovolo King the same plaudits. Chad Solomon is formerly of New York’s Milk & Honey, a bar that harvests huge ice blocks and breaks them down, and went on to co-found Cuffs & Buttons, a beverage consulting company. He said that his bar was Tovolo’s first commercial customer. King Ice Cube Tray is made of silicone that’s a step above what the FDA mandates for food grade material, and he also likes the chamfered edge. Asked why he prefers silicone to plastic or metal, Solomon said, “Plastic becomes brittle over time, and metal is not as flexible.” Although he highly recommends for barware, he says their silicone trays are a little more flexible, and can therefore be messier.

“I use the Tovolo silicone trays for nice square cubes at home,” says H. Joseph Ehrmann of San Francisco’s Elixir, designated as one of the best bars in America on multiple occasions by GQ, Esquire, and more. Kevin Liu, of and author of Craft Cocktails at Home, told us, “For the price, you can’t beat the 1-inch and 2-inch cube silicone trays by Tovolo.”

Our friend Lisa McManus is the senior editor responsible for all equipment testing for America’s Test Kitchen, Cook’s Illustrated and Cook’s Country magazines. With America’s Test Kitchen, she put King Ice Cube Tray through its paces, and came out really liking it. In her review, she said “the cubes look impressive, but more importantly, they kept our drinks from tasting watered-down. Over the course of 30 minutes, the Tovolo’s cubes only added 36 percent more liquid to our drink, less than any other ice cube tray we’ve tested.” In its roundup, the tray was Cook’s Illustrated’s most recommend among other large-cube models, beating out the company’s own smaller mold and Casabella’s tray.

On Amazon, Tovolo’s ice maker has a 4.5 star rating, with 303 5 star reviews. Of the negative ratings, it seems that some customers received knockoffs that imparted a chemical taste to the cubes. If that happens, Liu recommends “rinsing the cubes in ice water before using to get rid of the taste.” It’s worth noting that among our staff, we’ve never experienced such an issue. As for longevity, there’s no written warranty, but Tovolo’s Jennifer Carroll told us the company backs its products 100%.

Testing the Cubes

We ran our own tests, and found that larger cubes do indeed melt at a slightly slower rate, and smaller cubes at a higher rate. Some dilution is a good thing, because it “opens up the aromatics of the spirit,” as Creative Cocktail Consultant’s Van Flendern reminds us.

Our less scientific experiments found that the same volume of smaller ice does indeed make for a more watered-down taste.

No other ice cube tray we tested, of all our semi-finalists, cooled as moderately and consistently as the Tovolo.
No other ice cube tray we tested, of all our semi-finalists, melted as slowly as the Tovolo. It cooled a glass of room temp water by 26.8˚F in 10 minutes, diluting it by 26 grams. The other trays—Tovolo’s own Perfect Cube Silicone Ice Cube Tray and Casabella’s Silicone Big Ice Trays—made the water cool faster, which is less ideal, affecting the taste sooner with dilution between 27 and 28 grams. Not a ton but more than nothing. We also did some small cube tests to check that they were not better, to back up the theory we’d heard over and over again. (“Plastic trays are just ok. They work well enough, but as I mentioned when it comes to ice, bigger is better,” says Van Flendern.) Using the same 100 grams of ice but in two cubes instead of one giant one, the small ice cubes cooled the drink the most (28.5˚F) and also diluted the drink the most (28 grams). After publishing this piece, we had a smart writer, Dr. Drang, challenge the idea that using different sized ice cubes can make a difference in a drink. I think that’s a fair question. So we did another casual test specifically addressing this.

Is that enough to make a difference in taste?

I wanted to prove or disprove if using different ice cube sizes from our recommended Tovolo trays (big and small) would make a difference in taste.


So, I took an equal volume of ice but with different surface areas (four ~1+ inch cubes melted down a bit to match a single 2-inch cube). (Volume was measured by water displacement. And after I melted the small cubes a bit to get the volume to match, I stuck the cubes back in the freezer.) I put two ounces of Templeton Rye in a glass labelled it A. Did the same thing for glass B. After 10 minutes I removed the ice cubes and put them in the freezer to stop dilution and keep the testers from knowing which type of cube was in each glass. Three editors, tested blindly, presented with multiple tastes of each, could not tell a difference in taste after 10 minutes. I put the ice cubes back in. After 18 minutes I repeated the test, the same three editors could tell a difference, but not unanimously:

  • Two editors said glass “B”, which had smaller cubes, tasted more diluted. 
  • Another editor, who is not a morning person, said he could not tell the difference.

One of the editors who could taste the watered down drink also noted that they still tasted similar enough. After 25 minutes, the tastes were varied enough people could tell. This obviously isn’t a rigorous test, but it shows that using an ice cube in a shape with less surface area is noticeable, but still not a huge deal. In the end, if professional and award-winning bar tenders say use a bigger cube, we’re going to use a bigger cube.

Best small cube tray (for cooling water or juice or anything else quickly with a lot of dilution)

As far as small trays go for cooling water or other drinks where you don’t really care about dilution—warm soda or juice you just want to get to drinking temperature swiftly or to soften out the sugary taste—we have a pick, too.

Also Great
Tovolo's smaller set of trays gives you a total of 30 one-inch cubes, 15 in each. These are great for juice and water and other things you want to dilute more quickly. They're somewhat harder to get out than the King Cubes, but still a good option, especially if you have smaller glasses.
*At the time of publishing, the price was $10.
 We liked the Tovolo Perfect Cube Silicone Ice Cube Tray, which comes as a pair, each making 15 1.25 inch cubes at a time. It takes about three of these cubes to equal the same mass as one of the Kings, but the molds otherwise share the properties you should be looking for. If you have smaller glasses, these should be a consideration. Additionally, “smaller cubes will aid in keeping carbonated drinks fizzy,” says Sother Teague, beverage director at Amor Y Amargo, a bar that’s received such accolades as one of the best new bars of 2011.

Others we did not like so much. Gizmodo ran its own tests of ice cube trays of traditional style and size. Sam Biddle found OXO’s Good Grips Ice Cube Tray to be his favorite. It has a non spill design, but if you want that, we preferred their Good Grips NoSpill Ice Cube Tray. Its silicone lid allows it to be placed in a freezer at up to a 45° angle. Both make cubes with rounded bottoms, allowing them to be easily removed by pressing down on either end of the cube. Still, the Tovolo Perfect Cube Silicone was better.

Spherical cubes (a neat trick)

Neat but annoying
*At the time of publishing, the price was $9.
Although too finicky for every drink, we enjoy these sphere molds for giving cocktails and whiskies an orbital touch. They can be a little fragile if dropped when cold, however and need to be rinsed in warm water to be removed from their molds.
 The shape with the least surface area is the sphere; Tovolo makes a pretty neat set of Sphere Ice Molds. Each makes a 2.5 inch sphere, with a surface area of 19.6 square inches. This is with a higher mass than King Ice Cube Tray. Unfortunately you only get a pair of molds, and they’re more expensive—but man, do they look cool, and they fit in slightly smaller glasses. These are nifty—a few editors here have the King Ice as well as these spheres, and the feedback is that they’re expensive (that cost gets you two molds), more fragile (since they’re made from hard plastic in the lower portion) and because you have to rinse them in warm water to remove the cube, not the easiest to pop out. Still, all ice cubes are relatively cheap and these are a neat addition to a home bartender’s collection but definitely not beneficial beyond the wow factor.

Other Models We Looked At

There are dozens, if not hundreds of other models that I passed on, most of which are the generic, white, come-with-your freezer variety. There are very few large trays that match the qualities of Tovolo King. There are also lots and lots of novelty trays, but no one needs AK47 ammo shaped ice cubes. “Stay with cubes. All the funny/funky shapes (guitars, skulls, etc) have to many surfaces and just melt faster,” advises Teague.

In terms of more premium options, Cocktail Kingdom sells a Professional Ice Ball Maker similar to models coming out of Japan that cost hundreds of dollars. This one is a more reasonable $150, and makes 2.25 inch spheres by melting a cube down to shape between two heavy pieces of anodized aluminum. The result is a perfect sphere, but the cost isn’t worth it for home use; stick with Tovolo’s King, or splurge for the Sphere makers if you must.

Wrapping it up

Whether you’re shaking up cocktails or putting your whiskey on the rocks, Tovolo’s King Ice Cube Tray is the way to go. From the price, to the size and shape of the cubes it puts out, to the material it’s made of, it’s everything you should be looking for in an ice cube tray. It’s used in bars across the country, and there’s a good chance you’ve had its cubes in your drink without ever knowing the name. It’s an everyday luxury at a modest price.

October 9th 2013–Long term test results (flaws that are not dealbreakers)

These trays are holding up nicely for many readers and our own editors who own them. There have been no surprises, except two.

While we didn’t encounter either at first, after continued use, they’ve become more evident. We spoke to a representative from the company, who confirmed both. Not everyone will experience them, and neither is a dealbreaker, but they’re worth addressing.

The first is a white residue left behind on the tray or on the cubes themselves, which is attributable to hard water. As our contributor Kevin Liu explains in response to a reader comment, “Since the cubes in the Tovolo are larger, the water freezes more slowly and also circulates more within each cube before it freezes. This action deposits the minerals on the outside of each cube; with smaller cubes, the minerals get trapped inside the ice cube.” A similar sentiment was echoed by Tovolo’s Kerry Niesen when we asked her about the issue. “The only residue ever associated with these tests are the same elements that are in the water used to make the cubes. Essentially hard water residue that sticks to the ice trays and then to the ice cubes. They are more common in areas that add minerals to the drinking water.” If your water has a low percentage of dissolved solids, you may never see this residue, and if you do see them, there are no health concerns.

A more problematic issue is the smell the trays can develop over time. Some Amazon reviews describe it as “chemical,” while others simply say the Tovolo cubes stink. Reader franckhertz commented “Once I tried drinking the melted water from them and the smell became a poor taste.” As Niesen explains it, “Silicone is a rubber material and at low temperatures…can absorb odors from the freezer and refrigerator.”

Some workarounds

If either of these problems does pop up for you while using Tovolo’s King Ice Cube Tray, fear not. There are fixes, and they’re relatively simple. If a white residue does appear on the outside of the cubes, “simply rinse the cubes off with a little ice water (room temp water will cause your ice to crack) before depositing into your favorite drink,” says Liu. Many Amazon reviewers agree. We’ve also heard of some anecdotal evidence that using distilled or filtered water to reduce the deposits, although we don’t have hard science to back up this up.

A second fix can get rid of residue left behind on the tray itself, and eliminate foul odors, too. “The best solution is to clean your Tovolo ice trays in a dilute solution of vinegar and water about once per month,” says Niesen. “Sometimes it is necessary to soak the trays in the solution for up to an hour.” She recommends using equal parts vinegar and water. Additionally, storing the trays outside of the freezer when they’re not in use can help limit exposure to anything that may cause the smell. Niesen’s advice: “Once the cubes are frozen, decant them into an airtight container so they stay fresh in the freezer.” The reason for an airtight container? It’ll keep those smells from getting to your ice.

We haven’t experienced either issue to the degree that some people have, but we ran the cleaning regimen recommended by Tovolo. An hour after soaking in vinegar and water, our King Ice Cube Tray came out looking brand new, and with a noticeably diminished, although still present, freezer smell.

Appendix: the science of ice, melting, and chilling (as it pertains to drinks)

This appendix has been written by Matthew Nix, who has a PhD in biochemistry from UC Santa Cruz; and Kevin Liu, author of Craft Cocktails at Home: Offbeat Techniques, Contemporary Crowd-Pleasers, and Classics Hacked with Science, who also wrote a great article for Serious Eats called Cocktail Science: Five Myths About Ice, Debunked. Special thanks to Jim Meehan, award-winning bartender and author of The PDT Cocktail Book, for his thoughts on ice, too.

In our article on the best ice cube trays for home use, we recommended the Tovolo’s King Ice Cube Tray. Our recommendations were based on extensive interviews with master bartenders, aggregations of reviews from multiple sources and our own in-house testing. But what about the scientific justification for our recommendation? When we recommend something for your kitchen (the laboratory inside of every home), we like to try and cover all of our bases; that includes using sound scientific reasoning in addition to empirical evidence as to why one product is superior to another. We wanted to investigate even further into the science behind how different sizes and shapes of ice cubes melt and dilute your drinks. (And, of course, to find out why this does or does not matter in a drink.)

Here’s the trouble with trying to use science to explain reality: reality is complicated. In the classroom, we learn basic scientific principles by modeling ideal systems so the calculations are easier. The problem with this of course (according to the late statistician George E.P. Box) is that essentially all models are wrong. But some are useful.

…of all the science-y topics to examine, water and ice may be one of the most challenging.
And of all the science-y topics to examine, water and ice may be one of the most challenging. Consider that water is a powerful polar solvent; it is the only molecule to exist naturally on Earth in three different phases; it is less dense as a solid than as a liquid; and it forms the foundation for all physiological processes in the body.

It’s easy enough to pose the question “What type of ice melts the fastest?” But it’s much more complicated to ask “What type of ice from freezer temperature will melt fastest in a mixture of water and ethanol stored in a room-temperature glass held in a warm hand?” And, in the context of ice used to cool drinks, that is the only question that matters.

Ice cools drinks by melting

To really understand what happens when ice cools down your favorite scotch, you have to think a bit about the thermodynamics of heat transfer (and, in doing so, reach far back into the painful memories you might have associated with your freshman-year chemistry course).

First, what is heat transfer?  Heat and temperature are really just forms of kinetic energy, essentially a measure of how fast molecules are vibrating and moving. The transfer of that kinetic energy (aka heat) is spontaneous and flows from a warmer object (in this case your whiskey) to a colder object (the ice in your drink). This is also loosely known as the second law of thermodynamics.

When ice is dropped in your drink, heat begins to flow from the drink to the ice according to two general mechanisms, conduction and phase change.


Conduction is simply the transfer of heat between two objects that are in physical contact. How much heat is transferred, and the subsequent temperature change, depends on a characteristic unique to each substance called heat capacity.

The heat capacity of an object describes how much energy is required to change the temperature of that object by one degree. Water has a pretty high heat capacity: 4.186 Joules/gram*C (Joules being a standard unit of energy or heat, grams being…well, grams, and C being degrees Celsius). This means it takes more energy input to change the temperature of one gram of water than one gram of many other common substances—just one of its many amazing properties. Meanwhile, things like copper or iron have much lower heat capacity (0.385 and 0.452 Joules/gram*C, respectively) making them easier to heat up and useful for things like cooking.

Consider the simple (and delicious) case of a glass of scotch with some ice in it…
Consider the simple (and delicious) case of a glass of scotch with some ice in it (one of my favorite drinks, mind you). For didactic purposes, though, let’s also say that we are using some magical form of ice that will NOT melt (kind of like Ice-9) and can only exchange heat with our drink through conduction. What temperature will that drink be when it reaches thermal equilibrium (i.e., when the solid and liquid substances are equal temperatures)?

Ice has a heat capacity of 2.1 Joules/gram*C, which you’ll notice is lower than liquid water’s heat capacity. (This is because the water molecules in ice are arranged more rigidly in crystal form, which limits in their ability to move. The water molecules in ice absorb heat energy mostly through an increase in their vibrational energy, whereas the molecules in liquid water can absorb heat energy through an increase in vibrational AND kinetic energy, taking advantage of the kinetic energy of molecules that are able to zip around one another.)

Typical freezers will get ice down to -18°C, and scotch in my house is always at room temperature, around 20°C. Let’s assume we have equal volumes of solid and liquid, and that the heat capacity of scotch is that of pure water. (This is for simplicity’s sake; current theory is still struggling with calculating the heat capacity of ethanol-water mixtures, which, as it turns out, behave in extremely non-ideal fashions. See “excess heat capacity” for more on this.) With these assumptions, a quick back-of-the-envelope calculation (which scientists love to say) tells us that your scotch and Ice-9 drink will reach thermal equilibrium at 7.3°C.

Our fictional Ice-9 case is actually quite similar to what would happen if you were using whiskey stones instead of ice in your drink. Soapstone (which is what whiskey stones are made of) has roughly half the heat capacity of ice (0.98 Joules/ gram*C). So, utilizing conduction alone, ice has twice as much capability to cool down your drink as an equal mass of whiskey stones. However, since ice and whiskey stones have different densities (soapstone is 3 times as dense), comparing equal volumes of ice and soapstone means soapstone has slightly more capability to cool your drink down (just in our conduction-only model). Using the same conditions for our cocktail (but with stones instead of ice) your drink would get down to a similar 4.4°C. There’s not much of a difference, and it’s probably not noticeable to human perception.

Of course, while both ice and whiskey stones will cool down your drink to a certain degree solely through heat conduction, ice DOES in fact melt (which is good for all of us; see Kurt Vonnegut’s Cat’s Cradle). Which brings us to the other, more powerful mechanism for the transfer of heat from ice to drink: the phase change of solid ice to liquid water.

Phase change

Intuitively we all know that a phase change is the transformation of a solid to a liquid, or liquid to gas (as well as the reverse). To move from one phase to another requires a transfer of heat—either adding heat to melt solid ice, or subtracting heat to cool gas into liquid. What might not be obvious, though, the amount of energy that’s required to go from one state to another. For the melting of solid ice to liquid water, it turns out it takes a lot of energy, especially compared to how much energy it takes to simply change the temperature of said ice.

The energy needed to go from a solid to a liquid is captured by something called the specific heat of fusion. For ice, the heat of fusion is a whopping 334 Joules/gram. You might notice that unlike the heat capacity, there is no temperature component to this number. That’s because when you are adding energy to the ice for it to melt, there is no change in temperature of the ice—all you are doing is freeing molecules from a solid state to a liquid state (this is called latent heat, in case you were wondering). That ice you are melting was 0°C when it was solid (the melting point of ice), and after you put in a bunch of heat to make it liquid it’s still going to be 0°C.

(Technically the entire drink should should remain at 0°C until all the ice has melted. In general, any input of energy will go into melting more ice, not increasing the temperature of the liquid, at least if you discount temperature gradients within the liquid. (The liquid right at the ice-water interface will be 0°C, liquid up against the glass though might be slightly warmer; with mixing, though, it should be uniformly 0°C if there is still solid ice in the drink.)

Lets go back to our scotch and ice (real ice this time) example once more. After we drop that -18°C ice cube into our 20°C scotch, it will immediately begin exchanging heat with the scotch using both conduction and phase change. If THIS time we only consider the heat exchanged through melting (no heat exchange by conduction), how much ice has to melt to get our scotch down to 7.3°C, the lowest temperature our Ice-9 could achieve through conduction alone?

Since ice will melt at 0°C, lets also assume that our ice is actually at that temperature instead of a freezer’s -18°C. If your 0°C ice cube is 90 grams (pretty much like the Tovolo Ice King cube) and is dropped into 90 grams of scotch, it turns out you only need 14.3 grams of your ice cube to melt to drop the scotch to 7.3°C. That means you only need 16% of the ice cube to melt to achieve the same drink cooling that would be required of the entire cube through conduction. That’s almost an order of magnitude difference in ice’s ability to cool through its phase change rather than conduction alone. THAT’s what we mean when we say that ice cools drinks by melting.

It also highlights the enormous advantage ice has over a product like whiskey stones when it comes to cooling. Of course, that advantage does come at a cost. Whiskey stones will never dilute your drink; ice on the other hand will only do your cocktail justice for a limited time, after which your drink is too watered down.

Surface area of the cube matters when it is sitting in a drink

So we’ve seen that with ice in a drink, most of the cooling comes from the process of melting. However, it’s probably clear that it’s not just the total cooling capability of ice that is important for making your drink delicious. Beyond how cold you can get your drink, it’s also important to know the speed at which cooling occurs, since we are not waiting for the ice to completely melt before we take our first sip. That has to do more with the rate of cooling, which has a lot more to do with the ice’s surface area (and, thus, what the ice looks like).

The general equation to describe the rate of heat transfer includes an additional term that we haven’t seen yet: surface area.

rate of heat transfer = (U value) x (difference in heat) x (surface area exposed)

An examination of the above equation immediately tells us that we will get faster or slower melting (and hence faster or slower cooling) in direct proportion to how much surface area of the ice cube is exposed to the liquid. (This is also true of the “U value” or Overall Heat Transfer Coefficient which varies according to substance, but since we’re dealing with water it’s a constant in our case). The greater the surface area (think 90 grams of crushed ice) the faster the melting will occur. The smaller the surface area (a single 90 gram cube of ice) the slower the melting will occur. “So what?” you may ask. In the case of scotch, it makes a huge difference.

(Ice in a shaker is another story, where ice’s melting and cooling is accelerated by the physical agitation of the mixture and ice; in fact, in that case there appears to be little to no difference in the temperature and dilution results, regardless of what ice you use.)

Let’s assume that the scotch you are drinking tastes better with a little bit of water, but not too much water. For many, a little bit of water unlocks certain flavors and takes the edge off the booziness (there is some strong evidence that this actually does happen that’s not based on personal preference alone—see food science guru Harold McGee’s in-depth discussion regarding the effects of water dilution on the flavor profiles of myriad of drinks).

From McGee’s piece:

“Fans and judges of Scotch whiskies often sample their flavor by ‘nosing’ them, or sniffing the aroma that gathers in the glass. Nosers have long known that diluting the spirit with roughly the same amount of water reduces the alcohol burn. And at the same time, strangely, amplifies the aromas.

How can water reduce one sensation and amplify another? Both alcohol and aroma molecules are volatile, meaning they evaporate from foods and drinks and are carried by the air to the odor receptors high up in the nasal cavity.

Aroma molecules are also more chemically similar to alcohol molecules than they are to water, so they tend to cling to alcohol, and are quicker to evaporate out of a drink when there’s less alcohol to cling to.

This means that the more alcoholic a drink is, the more it cloisters its aroma molecules, and the less aroma it releases into the air. Add water and there’s less alcohol to irritate and burn, and more aroma release.”

It doesn’t take a lot of water to taste a difference in a drink, but it’s complicated. Jim Meehan, author of the excellent PDT cocktail book and award-winning bartender, said, “It’s very tough to test this objectively. I could probably taste a difference with as little as .5 oz. of water, which could happen very quickly with shitty ice. All very subjective though.”

(We are currently looking for a taste chemist to provide us with some objective data on this subject, and if we find something we’ll update this piece. Know anyone?)

So let’s assume that there is an ideal booze-to-water ratio where the taste of the drink is objectively perfect (for argument’s sake lets just say between 10:1 and 2:1, scotch to melt water by volume). When the drink has less water dilution (9:1 for instance) the drink tastes good. When it has more water dilution (2:1 perhaps, after more ice has melted) it tastes watered down. When your drink reaches that point, it’s past its prime and is never coming back. When you have ice with more surface area, the ratio of scotch to water changes much faster than when you have ice with less surface area. In essence, with a bigger cube your drink will spend more time in the optimal dilution range, giving you more time to enjoy your drink.

Up to this point we’ve only been spouting theory about heat conduction, surface area, optimal flavor profiles, and so on. But like we said, theory and models can only get you so far. When you are dealing with a complex system such as this one, eventually you just have to try things out and see what happens to confirm your theoretical hypothesis. So even though we’d already performed some in-house testing, we decided to conduct yet another real world test to better understand the melting, dilution and temperature changes that come with different kinds of ice. (Yup, we are officially obsessed.)

Here’s what we did

We started with equal masses of ice made in our home freezer (-15°C) using either the Tovolo Ice King Cube tray (one big 90 gram cube), the Tovolo Perfect Cube tray (three smaller cubes, 30 grams each for a total of 90 grams), or crushed ice (90 grams) giving us three test cases using the same mass of ice (and same volume of potential melt water, 90 mL) but markedly different surface areas.

Each kind of ice was then dropped into room temperature whiskey to start ice cube melting and drink cooling. (We specifically used 2 fluid ounces or 59 mL of whiskey at 68°F and a single Schott Zwiesel whiskey glass that we equilibrated in 70°F water prior to each trial.)

To simulate typical mixing and heating of the drink in a real world scenario, we hand-stirred the drink 4 times per minute. We weighed the solid ice and took the temperature of the liquid every 2 minutes for 30 minutes or until the ice completely melted, and then repeated the whole darn thing multiple times (a modest n=3) to gain at least a small amount of significance to our findings.

ice_massTo no surprise, the crushed ice melted the fastest. It was followed by the 3 cubes of small ice, and then, finally, the big ice. The difference in how much mass had melted between the small and big ice resulted in the big ice having a “lead” of around 4 minutes. (That is to say: it took the large ice 4 minutes longer than the small ice for either of them to reach a mass of, say, 40 grams.) So clearly the smaller surface area of the big ice is in fact slowing its melting rate relative to other forms of ice. So what does this mean when it comes to the actual dilution of your drink?

Graphing the data another way we get a picture of that crucial ratio of water and whiskey.

graph2-fixedWhat we see is that over time, different kinds of ice will dilute the whiskey faster or slower in direct proportion to how much surface area each kind of ice possesses. Although we have not performed taste testing to quantitate what constitutes the dilution threshold for when your drink becomes watered down, we arbitrarily set our whiskey drink “event horizon” at the 1:1 ratio of water to whiskey (see red line on the chart). Above this line, we think it’s safe to say the drink will for sure be watered down; below 1:1, your drink is more likely to still be intact. Given this line drawn in the sand, what’s the data telling us? For one, crushed ice ruins your drink the fastest. It arrives at the 1:1 ratio in ~8.5 minutes. Small ice arrives at the point of no return in 13.75 minutes, and big ice gets there in 18.5 minutes.

The take home message? Using big ice cubes will give you an additional 4½ minutes of time to enjoy your drink compared to smaller ice, and that time more than doubles when you compare it to crushed ice (if in fact the 1:1 dilution ratio marks the difference between a good and bad drink). That’s no small thing when you consider the typical time it takes to enjoy your drink in a bar with friends or relaxing at home; most people savor a good scotch drink and don’t chug it all that quickly.

Finally, the last thing we looked at during our tests was the temperature of the drink, shown below.

drink_tempA little unexpectedly, there was a very large difference between the absolute lowest temperature the drink was able to achieve and the type of ice used. For all cases, the biggest drop in the temperature of the drink occurred rapidly within the first 2-3 minutes. Crushed ice quickly and efficiently drops the temperature below freezing (recall that ethanol, like salt, depresses the freezing point of the entire drink) while the drink with big cubes never got below 0°C.

From this data, all ice appears to be about equal when it comes to the speed of chilling it down to the lowest temperature its going to get to, but vary significantly when it comes to just how cold they can make your drink. This raises yet another question about how different ice affects the bottom line. Specifically, how do different temperatures affect how much you enjoy your drink? Although it’s well known that colder temperatures have a numbing effect on the ability to detect flavor, determining what the optimal temperature is for cocktails is a bit tricky. (Though too far below freezing is probably too low to enjoy any kind of flavor.)

So, to wrap it up, there are both some scientifically sound reasons and empirical test results that make a good case for big ice. The smaller the ice, the better it is at cooling. And since most cooling comes from melting, small ice will also dilute your drink faster as we’ve found from our own testing. From this perspective then we have some pretty good reasons to use bigger ice. It will make your drink cold but not too cold, it’ll give your drink a bit of melt water but not too much, and finally, although it’s not scientifically sound and is just my humble opinion, big ice just plain looks cool.

To send this guide via email, fill out the fields below:
Message Sent!
Oops! Please try again


  1. Tovolo King Cube Ice Trays at Amazon
  2. Tovolo King Cube Ice Trays at Tovolo
  3. Jim Meehan & Joseph Schwartz, Making Ice, Speakeasy Cocktails: Learn from the Modern Mixologists
    "Large cubes slow dilution, allowing you to chill a drink (or keep a chilled drink cool) without it getting all watery. Use silicone molds to freeze flawless one-inch or two-inch squares"
  4. Lisa McManus, Chill Out with Tovolo’s King Cube Silicone Ice Cube Tray, America's Test Kitchen Feed, July 2, 2012
    "The cubes look impressive, but more importantly, they kept our drinks from tasting watered-down. Over the course of 30 minutes, the Tovolo’s cubes only added 36 percent more liquid to our drink, less than any other ice cube tray we’ve tested. Also, the bendable and pliable silicone makes popping out the cubes a cinch."
  5. Silicone Ice Cube Trays, Cook's Illustrated, September 1, 2011
    "The six massive 2-inch cubes look impressive, but more important, they keep drinks from tasting watered-down. Just be sure the cube fits the inner diameter of your favorite glassware— or they’ll get stuck halfway down the glass."
  6. Sam Biddle, The Best Ice Cube Tray of All Time, Gizmodo, May 26, 2012
    "This ain't smartphone and tablet territory, boyscouts—this is primal stuff. The spec that matters most here is the ease with which you can pry those precious cubes out of their womb and stick them into a glass before pouring it down your wide throat."
  • Christian

    Many of the reviews of the Tovolo on Amazon mention some “white residue” issue. If you’re worried about this: I’ve had a pair of Tovolo Kings for a while, and I have no idea what any of those reviewers are talking about. I’ve never had a problem.

    Don’t second guess these based on flaky Amazon reviewers; just buy them and start making better drinks.

  • Kevin Liu


    The reason some folks get a white residue with the 2″ trays is due to hard water. For example, my home tap water has a total dissolved solids level of over 300 ppm, which technically makes it a mineral water.

    Since the cubes in the Tovolo are larger, the water freezes more slowly and also circulates more within each cube before it freezes. This action deposits the minerals on the outside of each cube; with smaller cubes, the minerals get trapped inside the ice cube.

    To solve this problem, simply rinse the cubes off with a little ice water (room temp water will cause your ice to crack) before depositing into your favorite drink.

    • Chris Corwin

      I really like using distilled water to make cubes with in these trays.

      I have found that pouring boiling distilled water into them makes perfectly clear cubes, and doesn’t have any of the residue issues.

  • Rob Beschizza

    I have this one. Note that the cubes are beveled at the bottom. Otherwise great.

  • caseyljohnston

    Bought 2, thanks fellers!

  • sagehumphries

    I have these and they are great. No issues whatsoever!

  • KyleDeas

    My girlfriend gave me two of the spherical molds for Christmas, and I really love them – I find that rinsing them under the warm water is usually unnecessary and they melt at exactly the right rate. I haven’t tried the King Cube trays but I definitely wouldn’t qualify the spherical molds as annoying in the slightest, and I think the fears about their fragility are unnecessary.

  • SSteve

    Maker’s Mark sent spherical ice trays to its “ambassadors” as a Christmas gift in 2010 ( I love the ice spheres, but they are a little hard to get out of the mold. I usually have to run warm water over the mold to get them loose. I put the warmed, extracted spheres in a container in the freezer so I have cold ones ready to use. MoMA sells a similar mold that makes two spheres instead of four.

  • bregalad

    How can anyone recommend an ice cube tray without a lid? I don’t care how many packages of baking soda you put in your freezer, open trays produce foul tasting cubes. I see the trays don’t even stack either so you have to devote 3 times as much shelf space to ice cubes.

    • tfd2

      they actually stack pretty well. you just put one on top of the other. and if your freezer is that nasty, maybe you should toss the 3 year old corn dogs finally.

  • ryan harvey

    I noticed that nothing in the review noticed the smell of the tray after prolonged use. These trays make beautiful ice cubes, but beware that if you leave them in your freezer for a while, they really begin to stink! I pulled a tray out of my freezer after it had been sitting in there for a few months, and both the ice cubes and the tray smelled really bad with “freezer smell.” I washed the tray using several methods and nothing would remove the smell.

  • Daniel Harlow

    How about a picture of the ice cubes next time?

  • sockatume

    There’s a lot I could write here, but suffice to say I’m sceptical that your methodology for measuring cooling and dilution is sufficiently rigorous to measure the small variations you’re describing. The change is on the same order as the precision of your measurements.

    Blinding is also a concern.

  • Wiseguy

    They are two-inch wide cubes, right? “Size: 2-cubic inches” doesn’t sound very big. That’s 1.25″ x 1.25″ x 1.25″. Are we sure that isn’t two inches cubed (2″ x 2″ x 2″), which would be eight cubic inches?

    • Nick Guy

      Thanks! I’ve fixed that.

  • sal cangeloso

    Sounds like you (and Cooks Illustrated, etc) did your homework, but I had a set of silicone icecube trays in the past and they made my ice taste plastic-y. Maybe it’s not an issue with Tovolo, as I probably had an offbrand set.

    +1 on their Spherical Ice mold though. It’s an excellent product and very impressive to pull out if you are entertaining.


    Unorthodox, but we’ve found over the past 4+ years that the best combination of cube size (~2oz) and flavor control is the Beaba lidded tray. They think it’s for baby food, but we cocktail geeks know better. ;)

  • Glico Morinaga

    Bought one. Froze some water in it. Noticed absolutely no difference in ease-of-ejection from any other silicone tray I’ve had. These are every bit as annoying as all the others and the cubes don’t fit in most glasses. Whoopty doo.

  • Power My Life

    Not only are these ice cube trays great for not diluting your beverage, but they also help save on energy costs.

  • woberman

    I also had the “white residue” (which is much more disgusting that that phrase indicates…. for me, it looked similar to dissolved toilet paper you might see floating in a poorly maintained public restroom bowl). Rinsing the ice melt help, but my workaround was to throw away the Tovolo and buy a different ice cube tray.

    The white residue issue sounds rare, so you might ignore this issue safely. But I have other annoyances too: old school hard trays you can just bang against the ice bucket to cause 100% of the ice to fall out easily, but getting the ice out of the Tovolo was a one at a time & manual effort. Also, when filling a floppy silicone tray, it’s hard not to spill a lot of water in the process (filling, transporting to freezer, setting it down, etc…).

    I’m know I’m making a big deal out of minor complaints. But, I feel things that I interact with frequently should be simple and well designed. If the point is a special occasion large cube for drinks (which is what the review’s key issue), maybe the extra effort is worth it. For a day-to-day ice cube tray, I’d avoid this recommendation.

  • franckhertz

    I’m afraid I have both the smell issue that Ryan describes, and the white residue (presumably from hard water) that others describe. Washing them has proved hard to get rid of either the residue (left behind in the tray and is harder to get out than I would have thought). But the smell really lingers and is problematic. Once I tried drinking the melted water from them and the smell became a poor taste. :-/

  • Sigivald

    … plainly the ideal solution is to use a reverse griddle/peltier cooler for temperature, and add pure water for dilution.

  • Mike Riley

    I bought two of the Tovolo King cube trays based on this review and would discourage anyone else from making the same mistake. Within days they acquired a nasty smell that also translated into funky tasting ice and funky tasting drinks (which I unwittingly served to some guests… embarrassing).

    I tried cleaning the trays as described, but the smell/taste quickly came back. I’ve switched back to standard hard plastic trays, and while they’re not as easy to pop the cubes out of, they taste fine and don’t ruin the drink I’m trying to cool.

    Sweethome, I generally find your advice to be spot on, but foul tasting ice cubes that ruin a drink are a definite dealbreaker of a flaw… I think you should reconsider recommending this product.

    • tony kaye

      I’ll forward this along to our editors. Thanks for the feedback!

    • tony kaye

      Just got word back. We’ll be updating soon with non-silicone picks!