The Brita Longlast Filter can treat 120 gallons of water—three times more than most comparable filters—which lasts most families six months between replacements instead of the usual two. The filter also outperforms many of its competitors, with certifications to remove 10 contaminants, including 99-plus percent of lead (most competitors don’t have lead certification) and five so-called “emerging contaminants,” among them pharmaceuticals that have begun to enter the water supply. And it costs no more to use than the standard Brita filter, which isn’t nearly so well-certified.
The Longlast meets both of the primary NSF standards for gravity-fed water filters: Its Standard 42 certification means it improves taste by removing 97-plus percent of the chlorine compounds typically used to purify municipal water supplies, while its Standard 53 certification means it removes the vast majority of several contaminants—lead, mercury, cadmium, benzene, and asbestos. In addition, the Longlast is certified under Standard 401 to remove several “emerging contaminants,” including pharmaceuticals and pesticides. (See Our pick for details.)
Frustratingly, Brita does not currently package the Longlast Filter with any of its pitchers. But fortunately, the Longlast does fit almost all existing Brita pitchers as a drop-in replacement. And if you don’t already own a Brita pitcher, we have a pick.
The 10-cup Brita Everyday Pitcher is what most people envision when they think of a water filter pitcher. This affordable standard also happens to work with Brita’s Longlast Filter, so our two picks are easy to pair together (especially if, like many people, you already own an Everyday). Functional and easy to use, the Everyday has a narrow, space-saving footprint, as well as a lift-off lid that gives wide access to the upper reservoir, which has broad seamless surfaces that are easy to clean. In design details such as looks and ergonomics, some competitors have an edge on the Everyday, but with the Longlast installed, its filtration performance makes it exceptional.
The 10-cup Brita Grand Pitcher is another standard Brita pitcher that accepts the Longlast Filter, meaning it will produce water of identical quality. That’s the main reason we recommend it in the event you can’t find the Everyday, which we strongly prefer. The Grand does come in cheery colors that the only-white Everyday lacks, but it also has a drawback: The thumb-flip on its “easy-fill lid” (Brita’s words) is both slippery when wet—as easily happens when you’re filling the upper tank—and placed out of reach of anyone who doesn’t have big hands. It’s inconvenient to use your other hand to open the lid—you really need that other hand free to operate the faucet. The Everyday doesn’t have these problems.
Since 2016, Wirecutter writer-editor Tim Heffernan has overseen our water filter guides, covering both pitchers and under-sink systems. For this guide, he interviewed experts, helped design our tests, and lived with and used our past and current picks for months at a time. He is also the author of our air purifier and humidifier guides (among others), which gives him broad expertise in the arcane world of industry certifications and standards.
Co-author John Holecek has over 20 years of lab and field research experience in physics, chemistry, and earth and atmospheric sciences, including characterizing rainwater samples across the Pacific and Indian oceans (PDF). Working with Wirecutter since 2015, he has performed exhaustive, objective analyses of water purifiers, air filters, and humidifiers.
When we originally decided to test water filters, we also decided to limit our search to the most affordable and (not coincidentally) most popular option: filter pitchers. Since then, we’ve tested another popular type, under-sink filters that connect directly to the plumbing, but a filter pitcher is usually the first option for most people. And for good reason: Filter pitchers are inexpensive and easy to clean and maintain, and they can come with you when you move to a new home.
In our search, we focused on the simplest and most common type of filter pitcher: gravity-fed. You pour water into a top chamber, where it drains via gravity through the filter to a second chamber below. We also limited our search to filters that met NSF International’s certification requirements. NSF has two main testing standards that apply to pitcher-style water filters. Standard 42 measures the removal of chlorine taste and odor (municipal water supplies commonly add chlorine to kill pathogens), and Standard 53 measures the removal of numerous contaminants, including lead, mercury, benzene, and various biocides. NSF issues certifications for each contaminant individually, not for groups of contaminants, and few if any filters earn certification for everything.
We also sought filters certified to a fairly new NSF standard for water-treatment devices. Standard 401 measures the reduction of trace levels of emerging contaminants, including some pharmaceuticals, herbicides, pesticides, and manufacturing chemicals. The health effects of consuming trace amounts of these substances have not been established, and the EPA and other national health bodies do not regulate them or believe that their presence in water constitutes a public health threat. So we viewed Standard 401 certification as a bonus rather than a requirement.
Testing to ensure that a filter meets NSF standards is the responsibility of NSF or another accredited certification agency, such as the Water Quality Association. The WQA is a not-for-profit trade association that represents water-treatment manufacturers, suppliers, dealers, and “distributors of water quality improvement products and services.” The WQA is ANSI accredited to compete directly with NSF to ensure that water-treatment devices meet the NSF standards. This means that WQA certification possesses the same legal weight as NSF certification.
No filter completely removes all of a contaminant, even those contaminants for which the filter is certified. Certification means that a filter removes a significant percentage of a given contaminant—anywhere from 50 percent to nearly 100 percent—and reduces it to a level the EPA considers safe. And most filters do not remove “total dissolved solids”—primarily, minerals and organic compounds that exist in the water supply. That’s not a problem.1
Finally, using existing reports (particularly from Consumer Reports and sales data), trustworthy customer reviews, and our own experience, we came up with a list of eight filters (and their compatible pitchers) to test in 2016: two from Brita, one each from Mavea, Pur, Soma, and ZeroWater, and two from Clear2O (a faucet-fed model and a gravity-fed model). In 2017, we also considered a new filter and some compatible pitcher options from Brita.
For consistency in testing, and because it’s the most popular capacity, we chose the 10-cup (roughly 2.5-liter) pitcher from each manufacturer, or the closest size. Because manufacturers generally use the same filter cartridge across their whole line of pitchers, our test results are broadly representative.
In 2016, we evaluated each filter-pitcher combination on two subjective measures—taste and ease of use—and several objective ones, including speed of filtration, filter capacity, and, of course, the quality of the filtration itself. All our test pitchers had NSF 42 certification (for chlorine removal and taste), and most had several NSF 53 certifications (for various contaminants); we also conducted testing to see how they handled some chemicals that we believed would matter most to readers. In his own lab, John Holecek measured removal of chlorine, which has a big impact on water’s taste and smell, as well as each contender’s ability to remove “total dissolved solids”—basically, mineral salts and organic matter. For our top two picks, we contracted an independent test for lead removal, using a much more lead-contaminated solution than NSF calls for.2 (See the full report in Details from our 2016 lab tests.)
Our main takeaway from testing was the confidence that NSF certification is a broadly reliable measure of a filter’s performance. We found that filters performed well even when we tested them against far richer solutions of contaminants than you’d see in most US tap water. We learned that even some non-NSF-certified filters could perform exceedingly well on the certifications’ test requirements—for example, a filter that wasn’t even certified to remove lead topped 97 percent removal on a solution 16 times more concentrated than NSF calls for. If a noncertified filter can do that, we figure, any independently certified filters should perform exceptionally well.
This year, confident in NSF certifications, we went looking for potential replacements for our 2016 picks. (The Pur Classic, our top pick in 2016, developed a troubling habit of clogging for several owners, including a few Wirecutter staffers and Tim Heffernan, this guide’s co-author. Meanwhile, Mavea, maker of our runner-up filter pitcher in 2016, pulled out of North America.)
So we went looking for other options that we hadn’t tested before. In March 2017, Brita unveiled the Longlast Filter. Rated to 120 gallons or six months, three times the 40 gallons or two months of most pitcher filters, it immediately caught our eye. Its price (over the course of a year of use) was comparable to that of the standard filter. And it was certified by the Water Quality Association to NSF standards for chlorine, lead, cadmium, and some other contaminants. Given the lifespan, price, and certifications, as well as our 2016 observations of its competitors, there was no real chance that another filter could rival this option as long as it proved decent to use over time.
We called in a pair of Longlast filters and two compatible Brita pitchers, an Everyday and a Grand, and ran them through an extended real-world test. We evaluated the pitchers on their practical qualities: ease of use and cleaning, plus comfort in handling. And we monitored the filters to measure Brita’s stated claim of 120-gallon/six-month performance. Longevity in the lab is one thing—surviving in a 50-year-old New York City apartment with rusty pipes is another.
The Brita Longlast Filter is certified to remove lead, cadmium, and several other contaminants from tap water—including some pharmaceuticals, part of the so-called “emerging contaminants” that the EPA has identified as a growing concern. (This PDF lists the complete test results.) It also, more simply, produces great-tasting water. And it offers exceptional longevity: Virtually all pitcher filters are rated to filter 40 gallons of water, which equates to about two months’ use for a typical family of four. The Longlast is rated to filter 120 gallons, which means you’ll have to replace it only every six months. Finally, if you own a Brita pitcher already, it is very likely to be compatible with the Longlast—getting this improved performance is as simple as taking out the old filter and dropping in a Longlast.
Like every other filter we have tested, the Longlast is certified (in its case by the Water Quality Association) to meet the two main NSF standards for water filters. Its Standard 42 certification means it effectively eliminates chlorine, an addition to most municipal water supplies that acts as a disinfectant but can cause tap water to smell and taste “bleachy.” Its Standard 53 certifications mean it effectively eliminates five potential contaminants—specifically, more than 99 percent of lead, 95 percent of mercury, 97 percent of cadmium, 99 percent of asbestos, and 93 percent of benzene. That’s considerably broader performance than what you get from the standard Brita filter (which is certified only for copper, mercury, and cadmium), and in some ways better than our previous pick and runner-up, neither of which was lead-certified.
The Water Quality Association has also certified the Longlast Filter to meet the new NSF Standard 401 for emerging contaminants; specifically, it removes more than 90 percent of bisphenol A (BPA), estrone, ibuprofen, naproxen, and nonylphenol.3 That’s an uncommon distinction—few filters have yet to receive 401 certification.
Longevity also sets the Longlast distinctly apart from its competitors. Virtually all pitcher filters, including our past picks and the standard Brita filter, are rated to clean 40 gallons of water, or about two months’ worth for a family of four. The Longlast is rated to clean 120 gallons and should last a typical family six months between replacements. (The extended life is due to what Brita calls “pleated filter technology”—in essence, the designers folded up the filter, increasing its surface area without increasing its volume.)
In our real-world testing, a Longlast has now been in heavy use for three months—we refill the pitcher at least once a day—and has slowed down only slightly. (Filters work more slowly the longer you use them, due to sediment building up in the filter material.) Brand-new, our Longlast filtered a liter of water in 5 minutes, 10 seconds—exactly the same speed as our previous pick, the Pur filter. Now, it takes 6 minutes, 45 seconds. More practically, the brand-new Longlast filtered one complete fill of the upper reservoir—6.5 cups of water—in 7 minutes, 20 seconds. After three months of use, it now takes 9 minutes, 30 seconds. That feels like ages if you’re standing around waiting. But if you refill the upper reservoir each time you pour a drink, you’ll have a ready supply of filtered water whenever you’re thirsty.
The Longlast also costs virtually the same as the standard Brita filter: At this writing, one Longlast is about $15, while three standard Brita filters (that is, 120 gallons or six months’ worth) is about $14. And if you opt for a two-pack of Longlast filters, they sometimes cost less than the equivalent six-pack (one-year supply) of the standard Brita filter. So you get the benefit of less maintenance without additional spending.
Finally, the Longlast Filter will work in almost any Brita pitcher—the Stream and Infinity pitchers are the only exceptions. Otherwise, if your existing pitcher uses the standard Brita filter, the Longlast is a drop-in replacement.
Most owner reviews of the Longlast are strongly positive, regarding both its functionality and the taste of the water it produces. However, two complaints show up with some regularity: that the Longlast doesn’t seat properly in the pitcher, allowing unfiltered water to leak into the lower reservoir, and that the filter quickly clogs and stops filtering. We haven’t experienced either problem, but like anything, the Longlast isn’t flawless.
The 10-cup Everyday Pitcher is Brita’s simplest pitcher—boring, even—but we like it much better than the fancier Brita model we tested, the Grand. The Everyday’s lid comes off completely, making refilling easy; the Grand has a cumbersome “trap-door” refill opening that’s hard to operate with one hand, as it’s designed to be. The Everyday is also very neatly designed, with simple, flat surfaces everywhere—making it easy to clean inside and out. The Grand has a complex sleeve-tube in the lid, and also a battery pack and LCD screen, that collect grime and make cleaning difficult. Speaking of which: The Everyday doesn’t have an electronic refill reminder. Instead, you write the refill date on a sticker and affix it to the pitcher. This reminder method is simple, effective, and more reliable than the cheap electronics most fancier pitchers (of every brand) come with.
In past tests, we’ve found more comfortable pitchers to use—our previous pick, the Pur Classic, has a superb handle—but the Everyday is perfectly functional. And we like its narrow form, which doesn’t take up a ton of fridge space. Besides which, it’s compatible with the Longlast Filter—and that’s the most important thing.
It’s puzzling, and a bit frustrating, that Brita does not yet bundle the Longlast Filter with a pitcher as part of a package deal—you have to buy them separately. In addition, the upper reservoir of the Everyday Pitcher (and the Grand, for that matter) is unreasonably difficult to remove, which you need to do occasionally in order to clean the lower reservoir. You have to run your finger or thumb down the spout, grip very hard, and jiggle the reservoir while pulling up. There has to be a better way.
If the Everyday Pitcher is unavailable, the Brita Grand Pitcher is our reluctant choice as runner-up. We like the palette of bright colors it’s available in, and the fact that it accepts the Longlast Filter. But the thumb-operated “easy-fill lid” is anything but easy to use: It’s extremely slippery when wet (as easily happens when you’re filling), and it’s located so far from the handle that even people with big hands, like guide co-author Tim Heffernan and photographer Michael Hession, had trouble reaching it. (Wirecutter operations manager Sara Boyarsky could barely operate it at all.) That means you have to use both hands to fill the upper reservoir, when you really need one hand to operate the faucet.
The Grand comes with a little flap over the spout that’s designed to keep the pitcher completely sealed when it’s not in use. But it’s so annoying when pouring—it causes dribbles, and over time lime scale can clog its hinge and jam it shut—that we just removed it (see the photo). Many owners have reported problems with the lid falling off while pouring, too, though we had no trouble.
Finally, the lid has a complex internal shape that’s hard to clean (compare it against the simple lid of our pick, the Everyday, on the left), as well as a battery pack and screen that collect dirt in their crevices. And the Grand’s upper reservoir is a devil to remove when it’s time to clean the lower reservoir.
In a previous iteration of this guide, we chose the Pur Classic 11-Cup Pitcher (aka Pur LED) as the best water filter pitcher for most people. NSF International certified its filter to significantly reduce 10 contaminants from the water poured into it, including several heavy metals and biocides. In our test, the Pur filter performed better than or nearly as well as every other filter we considered in every category we thought was important, and it had no obvious flaws.
However, during long-term testing, several of our editors began to have problems with their Pur Classic and its Maxion filters. And many other Pur owners were running into the same problem. The filters were prone to “locking up” suddenly and no longer allowing water through—the result of air bubbles getting trapped in the carbon air filter particles. If you have experienced this problem, you should call Pur customer service at 800-787-5463, and the company “will make it right,” in the representative’s words. But none of us have had any luck with Pur’s suggested DIY fix (in the listed FAQs), and the company has backed out of its promise to permanently fix the problem with a new filter design. As a result, we no longer think the Pur Classic is the best water filter pitcher for most people.
We also recommended the Mavea Elemaris XL in 2016, as a slightly more ergonomic pitcher that performed about as well as the Pur Classic in our filtration tests. However, Mavea has since pulled out of North America, so we can no longer recommend its items. If you already own an Elemaris, be aware that no replacement filters will be available after the remaining inventory is sold.
In 2016 we tested the standard Brita filter (which comes with the Brita Everyday Pitcher we recommend). The standard filter left behind a significant amount of chlorine, earning a low score on water taste. And it is NSF 53 certified for only three contaminants (mercury, copper, and cadmium), whereas our pick, the Brita Longlast Filter, is certified to remove five (lead, mercury, cadmium, benzene, and asbestos), plus another five emerging contaminants under Standard 401. In short: You’ll get a standard filter with your Everyday Pitcher, but it’s best to ignore that one and install a Longlast instead.
Brita makes a wide range of other pitchers, and most of them work with the Longlast Filter. These include the Amalfi, Capri, Chrome, Classic, Marina, Metro, Mini Plus, Monterey, Pacifica, Slim, Soho, Space Saver, Stainless Steel, and Wave. We didn’t test them all; many are smaller than the 10-cup capacity we think is best for most people, and given that they all accept the Longlast, the only other differences are aesthetic. If you own one of these and like it, you have no reason to replace the pitcher—just the filter.
However, two Brita pitchers are incompatible with the Longlast Filter: The Stream and the Infinity. The Stream, it’s especially important to note, does not have any NSF 53 certifications. That is, it doesn’t remove any contaminants (such as lead or benzene, to take two examples). It simply removes chlorine—which water utilities add as a disinfectant—to improve taste, and as such it uses a completely different filter. (The Infiinity’s incompatibility comes down to a design issue.)
Clear2O made a name for itself with its CWS100AW, a highly regarded filter pitcher (Consumer Reports rated this model near the top of its list). This pitcher is unique in that it uses pressurized water from the faucet, rather than gravity, to force water through the filter. As a result, Clear2O uses a compressed block of activated charcoal rather than loose granules, and that results in truly spectacular NSF 53 certifications: 38 of them, including for lead, mercury, cadmium, and a host of organic compounds (biocides and petrochemicals for the most part). It also filtered a liter of water in just 17 seconds in our test, far and away the fastest result we got. So why don’t we recommend the CWS100AW? Because its faucet system is fussy—it was our least favorite pitcher to use—and involves permanently installing a special nozzle on your faucet and attaching it via a collar coupling to a hose on the pitcher every time you want fresh water. What’s more, the nozzle doesn’t fit on every faucet, and we couldn’t confirm that it fits on any pull-down faucet. Lastly, Clear2O is shifting its focus to a new line of gravity-fed filters, and though the company promises to always keep the CWS100AW (or something similar) on the roster, that’s a bit unnerving.
We also tested the new Clear2O gravity-fed pitcher, the GRP200. This model is NSF 42 certified for removal of chlorine taste and odor but not yet NSF 53 certified for any contaminant. It’s generally easy to use, but not spectacular or terrible in any aspect.
The Soma Pitcher is stylish but lacking in substance. Although the Water Quality Association certified it to meet NSF 42 for chlorine removal and taste, at the time of our tests it was not NSF 53 certified, which all the other filters we tested were. (Update: As of October 2016, the Soma filter is WQA-certified to NSF 53 for copper and mercury.) In use we found it tricky to pull apart for cleaning, and the wide spout made for problems when we needed to pour with precision (such as filling a drinking bottle). And despite meeting NSF 42 standards, the one-stage charcoal filter did the worst job of removing chlorine of any pitcher filter we tested.
The ZeroWater ZP-010 is NSF 53 certified to remove lead—the only other filter in our test with that distinction besides the Brita Longlast. However, it has a number of flaws, so we recommend it only if you really need that feature. We found it extremely difficult to use because the lid had a tight fit and no lip to push on, making the lid hard to remove for filling; instead, you must squeeze the tip of the lid tightly and rely on a friction grip to get it open. When trying to pour out the last cupful of water, we repeatedly had problems with the filter and fill housing falling out and onto our hands, the glass, or the counter. And despite the filter’s beefiness, it’s NSF 53 certified for only three substances (lead, chromium, and mercury), versus the Longlast’s five substances plus five more under Standard 401. The ZeroWater filter is also rated for a mere 15 gallons, versus 120 for the Longlast and 40 for all the others. And that fact only looks worse when you consider the current cost of a ZeroWater filter: about $15 apiece. Cleaning 120 gallons of water with a ZeroWater pitcher would require six filters and cost about $135; a single Longlast does the same for $15. Lastly, there are enough owner complaints about the ZeroWater model producing a “fishy smell” that the company addresses the matter on its FAQ page.
Tap water in the US is highly regulated by the Environmental Protection Agency and is generally of extremely high quality. The public water supplies are a critical resource and are carefully monitored, with thousands of tests performed each year for up to 40 or more compounds. Once a year, the public is required to be notified of the testing results via a Consumer Confidence Report. You can often find copies of these reports online via this tool, or see this example CCR (PDF) for a local municipal water supply in San Diego. The CCR details where your water comes from, how it is treated, and its test results. In 2014, for example, San Diego’s water supply was tested 6,779 times throughout the distribution system for total coliform and E. coli, with an average of 0.1 percent of samples testing positive (the range was 0.0 percent to 0.3 percent). Note that “testing positive” and “being dangerous” are not the same thing.
But even careful EPA testing is not the end of the story, as the Flint water crisis made plain. Water leaves municipal treatment plants in beautiful condition—but that doesn’t mean it arrives at your faucets that way. Lead pipes, leaded pipe solder, leaking water mains—those factors and more can introduce contaminants as water makes its way to and through your home. Quite literally, water quality can change from door to door, sometimes dramatically. So even though part of the CCR protocol involves testing the most at-risk end-user locations, it can’t necessarily tell you the quality of your water.
Bottom line: If you are concerned about the safety of your drinking water, whether due to lead contamination or other issues, it is important to have it tested. You could start with a simple test kit for less than $20 or collect samples and have them fully characterized by an EPA-certified laboratory for $200 to $300. If you own a well, the state of California put together a great guide (PDF) that tells how and when to test your water and what to test it for.
Lead was on everyone’s mind after Flint, and it is common in older plumbing—but “older” is younger than you might think. Only in June 1986 did the Safe Drinking Water Act declare that plumbing must be “lead free,” defined as solder and flux no greater than 0.2 percent lead and pipes and fittings no greater than 8 percent lead. (The 2011 Reduction of Lead in Drinking Water Act reduced this “wetted lead” from 8 percent to less than 0.25 percent.) But lead in plumbing does not equate to high levels of lead in the water—that’s a matter of various contributing factors like pH—and municipal water supplies are often doped with harmless chemicals that help “lock” the lead in the pipes and solder. (The issue in Flint was that these chemicals were removed, releasing the lead.)
The list of other potential contaminants is long. Some of the more common or familiar ones are mercury, cadmium, benzene, nitrates from fertilizers, and the generic herbicide 2,4-D (now in use for more than 70 years). At our request, Rick Andrew, NSF International’s director of global business development for water systems, prepared a very useful document (PDF) listing the chemicals that organization typically tests for when conducting Standard 42 and 53 certifications. This previously unavailable document details for each chemical the starting concentration and required ending concentration to pass the test. The required reduction can vary dramatically chemical by chemical—testing requires a 93.3 percent reduction for lead but only a 50 percent reduction in chlorine, for example.
In recent years NSF has identified some pharmaceuticals and other common compounds (BPA and DEET, for example) as “emerging contaminants” and has created a new test for them (Standard 401). The test is not common in water-filter testing yet, but may become so.
One last thing that is certainly in your water, but certainly not a problem, is the generic total dissolved solids. These substances are dissolved minerals and organic matter. According to the World Health Organization (PDF), “The principal constituents are usually calcium, magnesium, sodium, and potassium cations and carbonate, hydrogen carbonate, chloride, sulfate, and nitrate anions.” If many of those elements sound familiar, it’s because you can find a lot of them in a bottle of vitamins. The WHO further notes that studies indicate that higher TDS in drinking water is associated with lower incidence of cancer and heart disease. Besides, as the WHO says, “Water with extremely low concentrations of TDS may also be unacceptable because of its flat, insipid taste.” In short: At the levels typically found in drinking water, total dissolved solids are healthy and make the water taste good. ZeroWater, one of the pitchers in our test, makes a big point of reducing TDS to zero. That’s not self-evidently a good thing.
Most filter manufacturers give both a gallon rating and a suggested length of time between replacements. NSF test rules determine the gallon rating, but the replacement cycle is an estimate. Typically, 40 gallons of water filtered—the most common rating capacity—is assumed to equate to two months of usage. By the same token, the Brita Longlast Filter is certified to 120 gallons, with a suggested six-month replacement cycle. Of course, your usage may vary by a lot, depending on how many people use the pitcher, how much of a hydrator you are, and so on.
If you really want to be sure when your filter needs replacing, the only way is to measure your actual water usage. Try keeping track for a week and dividing the rated capacity (120 gallons for the Longlast) by the gallons of water you actually used. That’ll tell you how many weeks you should use each filter before replacing it. (This is a good time to note that pitchers with built-in replacement reminders—often a little light that turns from green to red—usually don’t measure your actual water usage. Most of them are simply clocks that measure the time elapsed since the last time you changed the filter.)
Not good at sticking to schedules? Here’s some peace of mind: NSF rules indicate that gravity-driven filters such as the Longlast undergo testing to Standards 42 and 53 only after they’ve already filtered twice the volume of water they’re rated for. The upshot: You can be pretty darn forgetful or lazy about replacing your filters and still be confident that you’re drinking clean water.
We evaluated each pitcher on two subjective measures—taste and ease of use—and several objective ones, including the speed of filtration, the filter capacity, and, of course, the quality of the filtration itself. Because all our test pitchers had NSF 42 certification and most had NSF 53 certification, we didn’t retest every single thing they could filter out. Instead, we limited our testing to a few factors we believed would matter most to readers. We focused on the removal of chlorine, which has a big impact on water’s taste and smell, and we looked at the ability to remove “total dissolved solids”—basically, mineral salts and organic matter. For our top two picks of 2016, we added an independent test for lead removal.
These pitchers typically use an activated-carbon filter, and some also use an ion-exchange filter. Briefly, an activated-carbon filter works by trapping pollutants in its high surface area via a chemical process called adsorption. Activated carbon is effective at adsorbing organic compounds, including taste compounds. Ion-exchange filters contain high-surface-area polymer beads that remove inorganic contaminants and heavy metal ions such as lead, cadmium, and mercury, and release small amounts of harmless sodium and potassium ions. You may remember sodium and potassium from biology class: The Na-K pump (Na = sodium, K = potassium) across nerve-cell walls creates the electrical nerve signal, so you need these ions to survive.
We used three water sources for our tests: tap water from the kitchen faucet in a 1960s San Diego ranch house (unfiltered water with the most contaminants) as a positive control, water from a fridge dispenser (in-line filter), and lab-grade reverse-osmosis-purified water (super-clean water) as a negative control.
Prior to testing we washed the pitchers by hand with dish soap, rinsed them, and dried them. While the pitchers were drying, we soaked the filters in cool tap water for 15 minutes and then flushed them under cool tap water for 10 seconds. Next, we installed the filters according to the manufacturers’ directions and filled the pitchers with water and discarded the water three times to remove any loose carbon dust. We then used water collected from the fourth fill for the following tests.
For the chlorine test, we tested water samples for temperature with a digital thermometer (roughly 20 degrees Celsius or 68 degrees Fahrenheit), then measured chlorine levels by inserting a test strip from Industrial Test Systems and color-matching the results according to the manufacturer’s directions. We also measured the samples’ pH using indicator strips, and found it to be neutral in all cases. We used an HM Digital TDS-EZ meter to measure total dissolved solids at the parts-per-million level (ppm).
Taste and ease of use are subjective measures, of course. Two testers tasted filtered samples from each pitcher two times and ranked them for overall good taste and lack of chlorine odor. We also ranked the pitchers’ usability on factors such as ease of filling and pouring, ease of replacing the filters and lids, and ergonomics.
|Model||TDS||Total chlorine (ppm)||Chlorine reduction (%)||Time to filter 1 liter (min:sec)||Subjective taste rank order (1 = best)||Usability (1 = best)||Overall rank|
|Lab reverse-osmosis (RO) water||31||0.0||100%||n/a||1||n/a||Excellent|
Summary of objective testing
For total dissolved solids, our tap-water control measured 380 ppm. One way to understand this measurement is with the analogy that one ppm is equal to one minute in two years (PDF). Most of our filtered water samples were close to this—at most 25 percent lower. The exceptions were the reverse-osmosis control and ZeroWater samples, which were near zero (31 ppm) and actually zero, respectively. However, ZeroWater specifically targets TDS, and the other filters do not. Moreover, dissolved solids are objectively not a bad thing and possibly an objectively good one—see the last paragraph of Contaminants that could be in your water for more details.
Total chlorine was high in the tap-water control (0.65 ppm; the EPA limit is 4.0 ppm). Water from the Soma (0.35 ppm), the Clear2O gravity (0.3 ppm), the fridge (0.2 ppm), the Clear2O faucet (0.175 ppm), and the standard Brita filter (0.1 ppm) left significant amounts of chlorine in the water after filtration. The Pur and Mavea, our 2016 pick and runner-up, reduced chlorine to zero or near zero (less than 0.03 ppm).
Comparing the taste rankings with the amount of time to filter reveals a pattern: Longer filtration times are associated with better taste, and shorter filtration times with worse taste. The Mavea, Pur, and ZeroWater filters all took more than five minutes to complete one filtration cycle—and they all tied for second in lack of chlorine taste, their water bested only by the lab-pure reverse-osmosis water. Soma, Clear2O gravity, and the standard Brita, by contrast, filtered in the one-to-two-minute range—and ranked a distant fifth through seventh place for taste. This makes sense: Longer filtration times mean more time for the filters to adsorb and absorb contaminants. It was interesting to see that principle so plainly illustrated in reality.
Lastly, but not shown in the table, is a useful ratio that we calculated for each pitcher: the volume of the top “fill tank” relative to the volume of the bottom “filtered tank.” That ratio varied widely, from a worst of 0.46 for the ZeroWater to a best of 0.73 for the Pur. In plain terms: You don’t get even half a pitcher of filtered water each time you fill the ZeroWater’s top tank, but you get almost three-quarters of a pitcher of filtered water every time you fill the Pur.
Based on our tests and on NSF certifications, we decided to conduct independent lead testing on our top two filters of 2016: the Pur Classic and the Mavea Elemaris XL, which both scored at or near the top on taste and ease of use and in our own tests of filtration performance.
The EPA’s actionable limit for lead in drinking water is 15 parts per billion, or 0.015 mg/L. Continuing the time analogy above, one ppb is equivalent to three seconds in 100 years. If a result is above that level in more than 10 percent of tests, the public receives notification and the operators of the water-treatment system must take a number of steps to improve the water’s quality. When NSF tests filters for lead removal, it begins with a lead level 10 times higher than that (0.15 mg/L) and requires that the filter reduce the level to below the EPA limit, to 0.010 mg/L.
Moreover, NSF standards require filters to maintain that performance for either 120 percent or 200 percent of their listed gallon capacity, depending on how the filter measures how much water has passed through it. For the Pur and Mavea, both rated at 40 gallons, that means they meet NSF Standard 53 after filtering 80 gallons—twice their rated capacity. (For more on how filters measure the amount of water that has passed through them, see How do you know when to change a filter?)
However (and happily), the municipal water in San Diego (PDF) has undetectable levels of lead, so in order to test the filters’ effectiveness, we had to raise our water’s lead levels artificially. Because we were testing only a single pass rather than a full life cycle, we also decided to start with a super-concentrated lead solution. We dissolved 10 mg of lead shavings (from a fishing sinker) with a 50-50 mix of white vinegar and hydrogen peroxide to form 40 milliliters of water-soluble lead acetate. We then mixed the lead acetate solution with tap water to make a 1-gallon stock solution with a target concentration more than 10 times above NSF’s 0.15 mg/L test solution.
EnviroMatrix Analytical, a local certified testing lab accredited by the California Department of Health Services Environmental Laboratory Accreditation Program (ELAP #1931), performed the tests for lead reduction using an inductively coupled plasma mass spectrometer (ICP-MS). We tested by filtering 0.75 liter of the lead-rich water through each pitcher, disposing of the contents into an aqueous-waste container, and then filtering a second 0.75 liter of water and using it for testing.
The results indicated that the control solution—our lead-doped tap water—had a lead concentration of 2.39 mg/L, or 2,390 ppb. In other words, we began our lead testing with a lead concentration 160 times higher than what the EPA considers unsafe and 16 times higher than the NSF test concentration. For comparison, 90 percent of homes in Flint, Michigan, had lead levels in their water of 27 ppb or less, though a team of researchers from Virginia Tech measured the highest level of lead in one home at 13,000 ppb.
Despite that heavy concentration, the Pur filter was able to reduce the lead levels by 97 percent to 0.073 mg/L. That amount is still seven times higher than the NSF Standard 53 requirement of 0.010 mg/L, but it’s way down from the highly elevated starting point. The Mavea filter substantially reduced the lead concentration as well—by 73 percent, down to 0.635 mg/L (60 times higher than NSF certification levels).
Again, the NSF Standard 53 test requires a 93.3 percent reduction of a 0.15 mg/L solution in order to pass—and the Pur filter achieved a significantly better percentage reduction than that on a water sample with 16 times more lead in it. However, given that NSF has not officially certified the Pur filter to remove lead and our test was at a single point (versus an ongoing test over the life of the filter), we cannot formally recommend it if you have a serious lead-contamination issue.