Protein Bars Tested: Heavy Metals, BPA, Acrylamide, and a Decade of Lawsuits (2026)

Save

We set out to find the cleanest protein bar on the market. What we found instead is a category where the Environmental Research Center has filed a Prop 65 lead case against at least thirteen brands since 2013, where Clean Label Project's independent lab testing found bisphenols in 15 percent of 165 bars and nearly every bar over California's acrylamide threshold, and where the two cleanest single sample lead readings in a 2026 screen both belong to brands with their own Prop 65 lead history on that exact product, one settled, one testing highest in the original case. This is not a simple story about which bar to buy. It is a story about why our own answer is to make one at home instead, and what the evidence shows for the rare times you don't.

1. What Is Actually in a Protein Bar

Before any lab result, the label itself is evidence. We read the ingredient list for every brand named in this article against our own hazard standard, which treats an undisclosed flavor umbrella, "natural flavors" or "natural flavor" with nothing named, as a disclosure failure on its own, the same way it treats an undisclosed fragrance in a cosmetic. It names no harmful substance, but it says the maker is choosing not to tell you what is actually in the bar.

Amrita is the case worth sitting with: the same brand names "chocolate flavor" specifically on one SKU and hides behind unnamed "natural flavors" on the next, which means this check has to be run flavor by flavor, never assumed from the brand's marketing.

2. The Real Packaging Finding: BPA and Phthalates, Measured

Our own materials standard scores packaging by what the contents can pull out of a plastic. A dry bar has little to dissolve a wrapper's additives into, which is why a theoretical read of a protein bar's flexible film wrapper would put it at the lowest risk tier, the same tier as a dry cinnamon jar. We said exactly that in an earlier version of this article, and it was too quick. A theory about migration is not the same thing as a measurement of it, and Clean Label Project actually measured it.

Clean Label Project's May 2025 study of 165 snack and nutrition bars tested every product for bisphenols and phthalates directly, not by inference from the wrapper material, and found detectable bisphenols in 15 percent of bars. The study attributes this to "leaching from food packaging and other materials" during manufacturing, including the kind of polycarbonate equipment and can linings used upstream of the final wrapper, not the shelf stable wrapper extracting into a finished dry bar over time. Phthalates were tested as a separate class for the same reason, contamination from "food packaging, processing equipment, and food preparation materials," though the published summary does not give an exact detection percentage.

Separately from contamination, there is a genuine waste finding. Every mainstream bar we checked, RXBAR, Clif, Quest, Built, ProMix, Rise, Aloha, 88 Acres, Zing, Amrita, ships in a multilayer laminate, usually a printed plastic film over a foil or metallized layer, sealed by heat, and almost none of it is accepted in curbside recycling because the layers cannot be separated by a materials recovery facility. A small number of brands outside the mainstream US market, Repaq based lines like Germany's The Nu Company and UK brand TRIBE, have built genuinely plastic free, home compostable wrappers for energy and protein bars. We could not find a single major protein bar brand sold on US shelves or Amazon doing the same.

3. Why Protein Bars Concentrate Heavy Metals

A protein bar is built from several ingredients that each carry their own, separate reason to pick up lead or cadmium, and a bar that uses more than one of them stacks the contributions rather than diluting them.

Put a cocoa coating, almonds, and dates in the same bar, a combination several best selling bars use, and you have stacked three separate soil uptake pathways into one product. A bar built from whey or collagen isolate with no nuts, cocoa, or dates has fewer places for a metal to ride in from, which is the main reason the lowest results in the testing below skew toward the simplest formulas, and why Clean Label Project's own data shows "kids" labeled bars, generally simpler formulas with fewer mix ins, testing lower across heavy metals, pesticides, and phthalates alike than their adult counterparts.

4. The Testing: 165 Bars, and a Closer Look at 25

Two independent datasets inform this article, and they are not the same kind of evidence. Clean Label Project is an established nonprofit watchdog with a public testing track record across multiple food categories; its May 2025 study screened 165 snack and nutrition bars from 50 brands through more than 20,000 individual lab tests run by Ellipse Analytics, an ISO accredited lab, for six contaminant classes at once: heavy metals, pesticides, glyphosate, bisphenols, phthalates, and acrylamide.

The organic finding is worth sitting with. Certified organic bars averaged 28 percent more heavy metals than conventional ones, and 6 percent of organic certified bars tested high enough in pesticides to exceed limits banned specifically from organic production, meaning the organic seal did not catch its own rule. Organic certification governs synthetic input use on the farm. It says nothing about what was already in the soil, which is where most of a bar's cadmium and lead actually come from, the same lesson our chocolate testing found.

Clean Label Project did not publish a full public ranking of all 165 products with individual numbers, but it did name a "Clean 16," the lowest contaminant load bars in the study. Five are genuine protein bars, the kind this article is actually about: Zing Plant-Based Protein Bar, 88 Acres Protein Bar, Amrita Protein Bar, Good To Go Keto Bar, and NUGO Organic Slim Bar. The other eleven are a different category entirely, fruit bars, kids' snacks, and low protein items wearing "protein" in the name without the protein to back it up: Autumn's Gold Grain-Free Bar, Bearded Brothers Bar, certain LÄRABAR flavors, That's It Mini Fruit Bar, Enjoy Life Protein Bites (6g protein, a "bite" rather than a bar, from an allergen free snack brand, not a protein brand), BoBo's Oat Bars Kids Line, GoMacro Kids MacroBar, Nature's Bakery Fig Bar Kids, Orgain Kids O-Bar, RXBAR Kids, and Skout Organic Kids Bar.

We originally checked only the three most obviously protein forward bars on that list, Zing, 88 Acres, and Amrita, and all three turned out to carry their own lead case. Pushed to look harder, we checked the remaining two real protein bars too: NuGo Nutrition (parent of NUGO Organic Slim Bar) settled its own Prop 65 lead and cadmium case in 2018, and Good To Go settled a separate one in 2021 for $55,000, though in both cases we could not confirm whether the specific Clean 16 SKU was one of the named flavors. All five protein bars on Clean Label Project's own best performing list carry a settled or unresolved Prop 65 lead case. The other eleven entries are not protein bars by this article's own standard, kids' snacks, fruit bars, and low protein items like Enjoy Life's Protein Bites, and we have not individually checked them. A Clean 16 placement describes relative contaminant load in one lab round. It is not the same claim as a clean legal record.

A smaller, newer, more heavily caveated screen

In 2026, the health app Oasis separately commissioned an ISO/IEC 17025 accredited laboratory to screen 25 retail protein bars for arsenic, cadmium, mercury, and lead by ICP-MS, one unit of each product. We want to be direct about what this is and is not. It is one sample per product, not a batch average; the lab is not named; and it is commissioned by a wellness app with its own commercial interest in the results, not an established watchdog with Clean Label Project's track record. We are treating every number below as a single data point, not a verdict.

With that caveat on record, here is the complete set of 25, worst to best on lead. We have converted the lab's micrograms per gram into the micrograms you actually swallow in one bar, because that is the unit California's limit is written against, and the conversion changes the ranking. A bar can look unremarkable per gram and still cross the line by being large, which is why serving weight sits in the table next to the result.

Source: Oasis, independent lab screen, samples received July 30, 2026, certificates issued August 18, 2026, one unit per product, ICP-MS via AOAC 2015.01 at an unnamed ISO/IEC 17025 accredited lab. Micrograms per bar are the lab's micrograms per gram multiplied by the labelled serving weight, benchmarked against the Prop 65 maximum allowable dose levels of 0.5 µg/day for lead and 4.1 µg/day for cadmium and the 10 µg/day arsenic no significant risk level. <LOQ means below the lab's limit of quantification, not zero. Every value above is a single sample, not a batch average.

Lead and cadmium do not move together here, the same pattern we found in chocolate. Rise Bar Almond Honey had the best lead result in the set but still carried measurable cadmium, and Blueprint Macadamia sat near the bottom on both while running the most arsenic per gram of any bar tested, plausibly from the macadamia nuts themselves. Only two bars crossed the cadmium limit, and Aloha owns the worst reading on each metal from two different products: the Peanut Butter Cup took the highest lead in the set at six times the limit, the Chocolate Chip Cookie Dough the highest cadmium at twice it. Aloha is also the one brand here facing an active, unresolved Prop 65 complaint.

The most useful rows in that table are the two bars built hardest against additives. Prima Cacao and Lineage Chocolate name every ingredient, refuse the phrase natural flavors outright, use grass fed tallow where most bars use seed oils, and in Lineage's case advertise third party purity testing with heavy metal testing on the cocoa specifically. Neither brand has a lawsuit or a recall we could find. Both still came back over California's lead limit per bar, at 137 and 130 percent, placing them mid table rather than at the top. A spotless label and a clean legal record did not produce a clean result.

The clearest line in the data is cocoa

Sorting those 25 bars by what is in them rather than by who makes them produces the sharpest split in this entire piece. We read every ingredient list and separated the bars containing cocoa solids, meaning cocoa powder, cocoa liquor, chocolate chips, unsweetened chocolate or alkali processed cocoa, from the bars containing none. Cocoa butter does not count here: it is the fat pressed out of the bean, and the metals stay with the solids, so Blueprint's white cocoa butter bar sits on the cocoa free side. Four bars of the 25 had no cocoa solids at all.

Every one of the 18 bars over California's lead limit contains cocoa solids. All four cocoa free bars came in under it, and the only two non detect lead results in the whole screen, Rise Bar Almond Honey and Clif Bar Crunchy Peanut Butter, are the two bars with no cocoa in any form. Serving weight does not explain it away either, because Clif Bar Crunchy Peanut Butter is the heaviest bar in the set at 68 grams and still read non detect, while much smaller chocolate bars crossed the line.

Cadmium moves the same way, which is what you would expect, because cadmium is the metal cocoa is best known for carrying. The median cocoa free bar held 0.86 µg against 2.23 µg for the median cocoa bar, only four of the 21 cocoa bars came in under the worst cocoa free one, and both bars that crossed the cadmium limit outright are chocolate. Arsenic then does the opposite, and this is the part that matters most. The two groups are indistinguishable on it, 0.83 µg against 0.66 µg at the median, no bar came near the limit, and the highest arsenic reading in the entire screen belongs to Blueprint Macadamia, a cocoa free bar.

That flat arsenic result is what makes the other two believable. If the cocoa free bars had simply looked better on all three metals, the likeliest explanation would be that cocoa free brands happen to be more careful across the board, which tells a shopper nothing about cocoa. Instead the gap appears on exactly the two metals cocoa is known to carry, cadmium drawn up from soil into the bean and lead settling on beans as they dry, and disappears on the one that arrives through rice protein, nuts and other plant inputs. An effect that is specific to particular metals is the signature of a real ingredient, not a tidier supply chain.

Cocoa is not an automatic disqualification, and we are not going to overstate a single sample screen. Three cocoa bars finished under the limit, and the lowest of them, Built Puff Coconut, carries cocoa only as a thin coating rather than through the body of the bar, which is the pattern you would expect if dose follows quantity. Lead also arrives through nuts and through plant proteins, which is the likelier driver in the two GoMacro bars built on sprouted brown rice protein. Only four of the 25 were cocoa free, so this is a strong signal from a small comparison group, not a settled rule.

For a reader standing in a shop, the practical version is short. Chocolate, cacao, brownie, cookie dough and anything described as a peanut butter cup all put cocoa solids in the bar, and that is where the lead and the cadmium concentrate, the same pattern the independent testing behind our chocolate guide turned up in cocoa itself. A vanilla, honey, coconut or plain nut bar is the better side of this particular bet. It is not a guarantee of a clean bar, because two of the four cocoa free bars here carry their own Prop 65 lead history, and it is the reason the recipe further down this page leaves the cocoa out.

5. Acrylamide: The Contaminant Almost Nobody Mentions

Every heavy metals headline in this category buries the single most consistent finding in Clean Label Project's data: 97 percent of the 165 bars tested exceeded California's Prop 65 threshold for acrylamide, a probable human carcinogen, the exact figure that drove a 2025 federal court ruling curbing Prop 65 acrylamide warnings as overly broad for a chemical this common in cooked food. Acrylamide forms through the Maillard reaction, the same browning chemistry that gives bread its crust and coffee its roast, whenever starches and sugars meet high heat. Roasted nuts, toasted oats, baked protein crisps, and browned chocolate coatings, the exact ingredients that make up most protein bars, are also the ingredients that generate it.

6. The Lawsuits: Why We Could Not Clear a Single Brand

Independent lab screens are useful and, as above, limited. What is not limited is a court filing. California's Prop 65 forces a different kind of disclosure: a private enforcer tests a product, files a 60 day notice, and either settles or litigates. We checked every brand with a low heavy metals reading or a Clean Label Project Clean 16 placement we could verify, and the pattern held across nearly all of them.

Every row above is from the California Attorney General's Prop 65 notice and settlement database or the Environmental Research Center's own settlement listings, unless otherwise sourced below. A notice with no public resolution found is recorded as unresolved, not assumed settled or dismissed.

Settled does not mean gone. Rise Bar, ProMix, 88 Acres, Zing Bars, and Blueprint each paid to resolve a lead case specifically because independent testing found lead in a named product, and a settlement is the one thing our own rules treat as a durable finding rather than a mere allegation. Rise Bar Almond Honey's non detect lead result in the 2026 screen above is the cleanest number in this entire article, and it is still not a pick, because the brand already admitted the underlying problem was real enough to settle over in 2024. A single later sample from an unnamed lab does not undo that.

Clif Bar's case is older, unresolved, and named the exact flavor. The Environmental Research Center sued Clif Bar & Co. in July 2013 over lead across its protein, energy, electrolyte, and snack bars broadly, and testing in that case found Crunchy Peanut Butter and Chocolate Chip the highest in lead of the flavors checked. We found no public record of how the case resolved. Crunchy Peanut Butter is also the exact flavor that came back non detect for lead in the 2026 single sample screen above, the best result in the set alongside Rise Bar. A thirteen year old, unresolved case that specifically flagged this flavor is not something one newer, unnamed-lab sample clears.

Unresolved is not the same as clean either. RXBAR and Amrita both have a filed Prop 65 notice with no public settlement, dismissal, or consent judgment we could find. We are not treating silence as exoneration.

Built Brands is a different problem entirely. Two separate 2024 federal class actions, Palacios and Sullivan v. Built Brands LLC, allege its Built and Puffs bars do not consistently deliver the 15 to 19 grams of protein stated on the label, with one plaintiff's independent testing reportedly finding as little as 8 grams in a bar labeled 17. That has nothing to do with heavy metals, but under our own rule a brand with an adverse finding on its own label claims cannot earn a pass on self reported evidence elsewhere either, and Built's Puff line also carries the undisclosed "Natural Flavors" line described above. An earlier version of this article named Built Puff Coconut as a "least bad option" with a buy link despite that disclosure failure. That was a mistake. The product did not pass our own scan, and we should not have featured it regardless of how it was labeled in the text.

Quest Nutrition: a pattern, not a finding on its bars

Quest Nutrition has drawn three separate rounds of heavy metal or ingredient safety scrutiny across three different product lines in four years: a 2022 notice on its Protein Chips, and a class action filed in early 2026 over lead in its ready to drink Protein Shakes and Milkshakes. We found no Prop 65 case naming a Quest protein bar specifically, so under our own scope rules we are not extending either finding to the bars, only noting the pattern across the rest of the brand.

7. Recalls: Allergens and Salmonella

Separate from heavy metals and lawsuits, protein and nutrition bars have a real, recent recall history, mostly undeclared allergens rather than chemistry.

• NuGo recalled its Dark Chocolate Chocolate Chip and Dark Chocolate Pretzel vegan bars in December 2024 for undeclared milk.
• Nutrabolics pulled all four flavors of its FEED ME Vegan protein bars in September 2024, also for undeclared milk.
• Wellness Natural USA expanded a recall of its SimplyProtein Peanut Butter Chocolate Crispy Bar for undeclared cashews.
• Gatorade Protein Peanut Butter Chocolate Bars were added to the Quaker Oats granola Salmonella recall in January 2024, which began with Quaker's own cereal and granola bars in December 2023.

None of these are open as of this writing, and under our own rules a closed, remedied recall over 24 months old is informational rather than a current finding; the Gatorade and Quaker recalls are approaching that line. We are listing them because undeclared allergens are a real, recurring failure mode in this category distinct from the chemistry above, and worth checking for on any bar with a mixed facility or co packer.

8. Is Any Protein Bar Actually Clean?

No. Clean Label Project found detectable heavy metals in all 165 bars it tested, 22 percent over the Prop 65 lead threshold and 97 percent over its acrylamide threshold. The smaller 2026 screen found 18 of 25 over the lead threshold and cadmium in every single one. Judged against the stricter 5 ppb ingested bar we hold every food, drink, and supplement to on this site, every detected result in either dataset clears that bar, because even a trace detection above the lab's limit of quantification is a finding under our own standard, not a pass.

Then there is the legal layer on top of the lab numbers, which is the part we underweighted the first time we wrote this article. We checked litigation history for every brand that was a realistic candidate, meaning it either led the 2026 screen or made Clean Label Project's Clean 16, thirteen brands in total, not just the ones an easy search turned up first. Thirteen for thirteen: Rise Bar, ProMix, 88 Acres, Zing Bars, Blueprint, NuGo Nutrition, Good To Go, and Clif Bar have all either settled or carry an unresolved case testing found highest in lead on the exact flavor in question; RXBAR, Perfect Bar, and Amrita have unresolved notices; Built Brands, the one brand with no metals case at all, has two separate 2024 lawsuits over its own label claims instead. We did not individually check litigation for every bar that simply appeared in the 25 bar screen with a mediocre or bad reading, among them Santa Cruz, David Protein, Daily Bar, Jacob, and GoMacro's Smooth Sanctuary flavor, since none of those were ever realistic candidates to begin with.

9. What Actually Reduces Your Risk

Your own favorite bar is likely not one of the brands named in this article. Type it into Product Check, on the site or in the app, for an instant verdict against the same independent lab tests, litigation records, and materials standard used throughout this piece.

10. Frequently Asked Questions