What are ultra-processed foods and how to recognize them?

1. There is no universal list valid for all recipes

A commercial category describes a use, not a composition. “Sandwich bread,” “fruit yogurt,” “plant-based burger,” and “soup” can refer either to a short recipe or to a formulation containing flavorings, texturizers, and food fractions. Any list that assigns every product with the same name to one group will misclassify some of them.

Nova’s authors instead propose identifying markers in the formulation. Their 2019 guide states that one practical way to identify Group 4 is to look for at least one substance rarely used in home cooking or an additive whose function is to recreate or intensify the product’s sensory qualities. Monteiro et al., 2019 — practical criteria for identifying Nova 4 foods

A useful list is therefore only a guide: it points to categories whose recipes deserve closer inspection. It cannot replace the label on the product you buy. Note the date as well, because reformulation can change the group without changing the product name, package design, or brand.

2. “Processed” does not mean “ultra-processed”

Cooking, drying, fermenting, pasteurizing, freezing, and canning all process food. These operations can improve microbiological safety, digestibility, shelf life, and year-round access. Nova still places many foods that undergo these processes in Groups 1 or 3.

Plain frozen vegetables typically remain minimally processed. Tomatoes canned with salt may be classified as processed without being ultra-processed. Dried pasta made from semolina and water does not become Nova 4 simply because a factory kneaded and dried it.

ANSES emphasizes that the term “ultra-processed” can be confusing because Nova Group 4 relies more on formulation than on a precise measure of processing intensity. This distinction explains why “industrially manufactured” and “Nova 4” are not perfect synonyms. ANSES — opinion on so-called ultra-processed foods

• Understand Nova’s four groups and history →

3. Common Nova 4 examples are starting points, not verdicts

The easiest formulations to identify often combine a fractionated base, several texturing or flavoring agents, and a ready-to-eat format. The examples below commonly appear in Nova descriptions, but every row comes with one condition: check the actual recipe.

4. First look for substances rarely used in home cooking

Some ingredients indicate that a manufacturer separated raw materials into fractions and then recombined them. Common examples include maltodextrin, hydrolyzed proteins, certain protein isolates, invert sugar, glucose-fructose syrup, and hydrogenated or interesterified oils. No single Nova 4 product necessarily contains all of them.

In this context, their presence helps classify a formulation; it does not declare the substance toxic at the amount consumed. Regulatory safety, nutritional contribution, and a substance’s role in Nova are three different questions. “Rare in a home kitchen” does not mean “poison.”

Conversely, home cooks may use less familiar ingredients depending on culture, country, or medical needs. The kitchen-familiarity test therefore has a contextual element. When classification informs research or policy, a more precise protocol is required.

• Identify fractions and derivatives that replace an intact food.
• Distinguish a culinary ingredient from a reconstructed combination.
• Do not infer danger from a chemical or technical name.
• Research the function of the ingredient before concluding.

5. Additives provide different clues depending on their function

Nova emphasizes so-called cosmetic additives: colors, flavorings, flavor enhancers, sweeteners, texturizers, and other substances used to imitate, mask, or intensify sensory properties. They suggest a Nova 4 formulation more strongly than an additive used only to preserve or protect the food.

But no reliable rule says “one E-number means ultra-processed.” Some Group 3 foods may contain a preservative or antioxidant. Conversely, a formulated product may belong to Group 4 because of a fractionated substance even when the word “additive” is not obvious to the reader.

European labels identify the functional category and the name or E-number of many additives. Use that information to understand an ingredient’s role in the recipe. It does not reveal the amount, the complete industrial process, or an individual risk estimate by itself.

6. A four-step method works better than the five-ingredient rule

Step 1: identify the exact product, brand, recipe, and date. An old photo or database entry may describe a different formulation. Step 2: identify the starting food and simple culinary ingredients. If they fully explain the product, it is more likely to belong to Groups 1–3.

Step 3: look for fractions rarely used in home cooking and additives with sensory functions. Do not merely count ingredients. A salad with eight vegetables and a simple dressing may contain more than five ingredients; a four-ingredient drink may consist of water, sweetener, flavoring, and coloring.

Step 4: rate your certainty. “Probably Nova 4” differs from “confirmed by an explicit recipe and documented rules.” If the list is missing, generic, or ambiguous, keep the classification “undetermined.” This honesty avoids fabricated data and helps compare the same products over time.

• Product and recipe precisely identified.
• Recognizable food base and culinary ingredients identified.
• Formulation markers and additive functions examined.
• Classification recorded as certain, probable, or undetermined.

7. Bread, yogurt, and high-protein products illustrate the gray areas

Bread made simply from flour, water, yeast or sourdough, and salt typically belongs to Group 3. Packaged bread containing isolates, emulsifiers, flavorings, and colors may fall into Group 4. Packaging is not the criterion; the recipe and purpose of the ingredients are what change.

Plain yogurt made from milk and cultures differs from a dairy dessert reconstructed with flavorings, sweeteners, modified starches, and texturizers. Between those examples, certain sweetened or thickened recipes create genuine classification disagreements. Acknowledge uncertainty instead of inventing a clear boundary.

“High protein” does not protect a product from Group 4 either. Many protein bars, drinks, and desserts use isolates, flavorings, and texturizers. That does not make them useless: a convenient product can help meet intake needs in some situations. Evaluate the recipe, calories, protein, tolerance, price, and what the product replaces at the same time.

• Evaluate high-protein products without relying on their marketing halo →

8. Reducing ultra-processed foods does not mean banning every convenient product

A sustainable diet is built around choices you can repeat. Ready-to-eat foods can provide convenience, safety, preservation, fortification, or accessibility. They may be especially important when time, budget, disability, housing, work, or medical needs limit cooking.

The useful goal is not to achieve one “zero Nova 4” day and then give up. Identify products eaten frequently and in substantial amounts, especially when they displace fruits, vegetables, legumes, whole grains, or suitable protein sources. Start with the product whose alternative is genuinely available and acceptable.

Frequency matters more than simple presence. An occasional dessert within a varied diet does not carry the same weight as a sugary drink, snacks, and several ultra-processed meals every day. Do not moralize a food—or a person—based on one package.

• See what studies actually say about health risk →

9. Run a seven-day audit before changing your diet

During an ordinary week, record each recurring product, its serving, and the context: a rushed breakfast, work snack, evening meal, or drink. Assign a definite, probable, or undetermined Nova status. This review shows whether a few products account for most exposure or whether industrial formulations appear throughout the day.

Then rank possible changes by impact and ease. Replacing a daily beverage can remove several weekly servings without adding meal preparation. Cooking two extra servings one evening can replace two ready-to-eat meals. A swap that sharply raises cost or removes a favorite food may not last.

Keep energy and macros in the analysis. Replacing an ultra-processed, protein-rich option with a minimally processed meal that no longer meets your needs can create another problem. Conversely, a calorie-dense “natural” option does not automatically support fat loss.

• Observe seven days without seeking perfection.
• Count recurring servings, not only different products.
• Choose a dominant source and a concrete alternative.
• Test the replacement one to two weeks.
• Keep the change only if it remains compatible with hunger, budget, time, and your goal.

10. Remember: read the ingredients, then place the product back in the meal

The most useful clues are fractionated food substances rarely used in home cooking and additives that reconstruct sensory qualities. Ingredient-count alone, factory production, packaging, an organic claim, or an app number is never sufficient in isolation.

Classification is only the beginning of the decision: what serving, how often, what role in the diet, and which alternative? This second step prevents Nova from replacing nutrition fundamentals. The goal is a mostly nutritious, feasible diet—not perfect control over every ingredient.