News

The "Non-UPF" Label Arrives. The Science Behind It Is Still Catching Up

Editor

17 Oct 2026 • 6 min read

Oct 17, 2026

The Non-GMO Project, the nonprofit behind the butterfly seal that reshaped how Americans think about genetically engineered food, has a new target: ultra-processed foods. Its Non-UPF Verified standard was first published in November 2025 (Food Business News), and by February 2026 brands such as Simple Mills had products verified under it (Business Wire). More than 300 products across 23 brands now carry the seal, including items from Simple Mills, Chomps and Blue Diamond (Food Dive).

The pitch is simple. Consumers are worried about ultra-processed foods, and they don't know how to spot them. A seal is meant to solve that. CEO Megan Westgate said the organization launched the program because people "were asking deeper questions about how their food is made" (PR Newswire).

But the reality is messier. There is no agreed definition of "ultra-processed" in the United States (CSPI). The science linking processing itself, independent of nutrition, to poor health is suggestive but not settled. And the certification market is already crowded with competing standards, each measuring something slightly different.

What the Non-UPF Verified seal actually tells a shopper depends entirely on what you think the problem with ultra-processed food is.

What the label actually measures

The standard rests on two dimensions: ingredient integrity and formulation, and processing limits (CSPI). It prohibits nearly 300 ingredients and bans non-nutritive sweeteners (CSPI). At least 70% of a product's ingredients must be minimally processed (Food Business News).

Added sugar is capped by category, from 2% or less by weight for snack foods and savory sauces to 40% for confectionery (Food Business News).

Westgate has acknowledged that some categories are hard to fit into a single rule. Baking mixes, she told Food Dive, are not eaten as they come out of the package; the consumer adds other ingredients first (Food Dive). The example illustrates the core difficulty: a single ingredient blacklist or sugar threshold cannot classify every food.

The standard is also more permissive than at least one competitor. Extrusion, a process often cited as a hallmark of ultra-processing and used to make breakfast cereals and corn chips, is not prohibited. High-moisture extrusion is permitted outright, and high-heat extrusion is "conditional"; conditional methods can make up as much as 30% of a product's formulation (CSPI). The rival Non-UPF Program, run by a different nonprofit, excludes extrusion entirely (CSPI). A consumer who assumes "Non-UPF Verified" means "no extrusion" would be wrong.

The evidence problem

The case for worrying about processing, as distinct from nutrition, rests largely on two studies that supporters cite often.

The first is a 2019 NIH experiment led by Kevin Hall and published in Cell Metabolism. Twenty adults lived in a metabolic ward for four weeks. In random order, each ate an ultra-processed diet for two weeks and a minimally processed diet for two weeks. The meals offered were matched for calories, sugar, fat, fiber and sodium, and participants could eat as much as they wanted (Healio). During the ultra-processed phase, participants consumed about 500 more calories per day and gained roughly 0.9 kg (Healio). They also ate faster, and during the minimally processed phase they lost about the same amount of weight (CORDIS).

The study is striking. It is also small and short. And the design could not isolate what drove the overeating: Hall has said his studies were not built to separate the effects of energy density (calories per bite) from hyperpalatability (KRDO/CNN). Headlines were less careful; NPR's 2019 story presented the finding as showing that ultra-processed foods drive weight gain (NPR).

The second is a 2024 umbrella review in The BMJ, drawing on 45 pooled analyses covering nearly 9.9 million participants. It found associations between higher UPF consumption and 32 adverse health outcomes, including cardiovascular death, type 2 diabetes, obesity, anxiety and depression (Cardiovascular Business). But every analysis it reviewed came from observational studies (The BMJ), which can show association but not causation. Outside experts also noted that few of the underlying studies were designed specifically to measure ultra-processed food intake (Scimex).

None of this means processing is irrelevant. It means the science is not yet precise enough to justify a binary label that tells consumers a product is definitively good or bad on this dimension alone.

A certification boom, with no referee

The Non-UPF Verified seal is one of three non-UPF certifications now on the market (CSPI).

The Non-UPF Program certifies products that fall into NOVA Groups 1 through 3 and excludes anything in Group 4, the classification system's ultra-processed category (CSPI). WISEcode, a food-scanning app company, launched a verification program called the Non-UPF Shield in early March (FoodNavigator). It sorts products into five processing levels from Minimal to Super-Ultra, and only the lowest three qualify (CSPI). Confusingly, WISEcode's mark also reads "Non-UPF Verified," the same name as the Non-GMO Project's seal (CSPI).

The Center for Science in the Public Interest compared the three and found they differ in how they define ultra-processing and in which processes, ingredients and nutrients they weigh. Its verdict: because each label means something different, the labels taken together are not much help to shoppers. CSPI also judged that the Non-GMO Project's standard likely admits a broader range of products than the Non-UPF Program's (CSPI).

The Non-GMO Project's standard carries the most commercial weight. The organization has been verifying products since 2007 (Food Business News), and its butterfly seal appears on more than 63,000 of them (Business Wire). That scale is an asset for the new label. It is also a reason to ask harder questions about it. Brands are already marketing hard on the seal: Simple Mills ran a national ad campaign in August around its certification, although only about 60% of the brand's products were certified at the time (CSPI).

Governments are moving, slowly. The FDA and USDA issued a request for information on a federal definition in July 2025 (Federal Register). On August 10, 2026, HHS and USDA submitted a proposed definition for final review (Quality Smart Solutions), framed as a white paper rather than a regulation (FoodNavigator). As of October 13, its text had not been made public (National Law Review). California has gone further: Gov. Gavin Newsom signed a law creating a voluntary, state-backed non-ultra-processed seal, though it is not expected on shelves before mid-2029, and the Non-GMO Project's program opposed the bill (New Hope). In Congress, a bill introduced in July by Sens. Richard Blumenthal and Cory Booker would direct the FDA to set a single official definition of "non-UPF" (CSPI).

Until any of that takes effect, private certifiers fill the vacuum, each with its own definition and its own commercial interest in expanding its seal.

What the label doesn't tell you

Three things are worth keeping in mind.

Non-UPF does not mean healthy. The standard does not evaluate a product's nutrient content or overall nutritional adequacy (Food Business News). Of the three non-UPF seals, none considers sodium or saturated fat, and only two look at added sugar (CSPI). A product can be minimally processed and still high in saturated fat or sodium. Certification complements a look at the nutrition panel; it does not replace it.

The certifier has a stake. The Non-GMO Project is a mission-driven nonprofit, but extending its certification model into a new category also expands the market for its seal. Certification is typically paid for by the companies seeking it, which creates a standing incentive toward standards loose enough to certify more products (CSPI). That doesn't make this standard wrong. It does mean its claims should be read the way you'd read any certifier's claims.

Accessibility is the elephant in the aisle. Ultra-processed foods are often convenient, affordable and shelf-stable. Even Hall, whose research is central to the case against them, has stressed that cooking less-processed food takes more time and more money (CORDIS). A certification that helps affluent shoppers feel better about their grocery cart while leaving the underlying food system unchanged is not a public health intervention. It is a market segmentation tool.

The bottom line

The Non-UPF Verified label is a genuine attempt to address a real gap: consumers want to know how their food is made, and current labels don't tell them. The organization behind it has a long track record.

But the label arrives before the science has settled, ahead of a federal definition, and alongside competing seals, one of which shares its name. It measures something real, processing intensity and ingredient quality, but it does not measure health. It is one data point, not a verdict.

For now, the most honest thing a consumer can do is treat the seal as a signal that a product has cleared a particular bar set by a particular organization with a particular worldview. That is useful information. It is not the same as knowing the food is good for you. (ep)