What is the Nova classification, and what are its four groups for?

1. Nova classifies processing; it is not a health score

Nova’s authors define it as a classification of foods according to the extent and purpose of processing rather than solely by nutrient content. They also clarify that “Nova” is a name, not an abbreviation. Each food is assigned to one of four descriptive groups. Monteiro et al., 2016 — formalization of Nova’s four groups, Monteiro et al., 2019 — definition and identification of ultra-processed foods

The number should not be read as a score out of four. Group 2 includes salt, sugar, butter, and oils: ingredients used in cooking, not a middle category that should necessarily be eaten more often than Group 3 foods. The classification describes a product’s role in manufacturing and meals.

That distinction changes how Nova should be used. It can prompt you to look beyond calories and distinguish a recognizable food from a ready-to-eat formulation. It cannot determine by itself whether a serving fits your goal, whether a product meets protein needs, or whether your diet provides enough fiber and micronutrients.

2. The story began in 2009 with a different way of examining the food supply

In 2009, Carlos Monteiro published a commentary calling on public-health nutrition to pay greater attention to industrial processing. The initial proposal included three groups and already emphasized that processing is not inherently harmful: cleaning, cooking, fermentation, and preservation can be necessary and useful. Monteiro, 2009 — foundational commentary on processing and health

A 2010 publication then applied a classification based on the extent and purpose of processing to food purchases in Brazil. The framework continued to evolve. In 2014, Brazil’s official dietary guidelines built recommendations around fresh or minimally processed foods, cooking, and limiting ultra-processed products. Monteiro et al., 2010 — early application to food purchases in Brazil, Brazilian Ministry of Health — 2014 national dietary guidelines

The four-group version was formalized in the mid-2010s and further detailed in a 2019 identification guide. Nova did not begin as a commercial logo calculated by a brand; it began as a framework for research and food policy. Its revisions also explain why any definition should be tied to a version and coding method.

3. The four groups describe four different functions

The classification becomes clearer once you stop merely counting factory steps. Pasteurized milk can remain in Group 1 despite substantial industrial processing, while a drink reconstructed from food fractions and flavorings typically belongs to Group 4. Purpose, formulation, and the role of ingredients matter as much as the number of operations.

4. Group 1 includes processes that preserve the original food

Group 1 includes edible parts of plants, animals, and fungi, as well as water and minimally processed versions of those foods. Removing inedible parts, washing, portioning, drying, grinding, refrigerating, freezing, pasteurizing, or vacuum-packing does not automatically move a food out of this group.

This definition avoids equating all manufacturing with nutritional degradation. Plain frozen vegetables, roasted coffee, rolled grains, and pasteurized milk have undergone processing, sometimes using advanced technology. They still retain the food’s central identity and have not received salt, sugar, or fat to become a prepared product.

The category has a practical benefit: it puts foods and cooking at the center. Its limitation is that two Group 1 foods can have very different nutritional profiles. Water, a banana, and fresh meat do not share the same energy density or nutrients; Nova does not claim they are nutritionally equivalent.

5. Groups 2 and 3 show why not all processing is “ultra”

Group 2 contains processed culinary ingredients obtained from Group 1 foods or from nature. Oils, butter, sugar, and salt are classic examples. Their purpose is to season, cook, or prepare a dish. They are normally used with other foods, and the amount still matters nutritionally.

Group 3 includes relatively simple products made by adding a Group 2 ingredient to a Group 1 food, often to preserve it or make it more palatable. Vegetables preserved in brine, salted nuts, certain cheeses, and bread with a simple recipe typically fit this logic.

These groups correct a common shortcut: canning, fermentation, or breadmaking does not make a food Nova 4 by definition. The same commercial name can still cover very different recipes. Bread made from flour, water, sourdough or yeast, and salt is not classified like packaged bread containing emulsifiers, flavorings, and fractionated substances. You must read the actual ingredient list.

6. Group 4 is identified mainly through formulation and industrial markers

The 2019 identification guide describes ultra-processed foods as industrial formulations often made from numerous substances. Markers include fractions rarely used in home cooking, such as certain hydrolyzed or isolated proteins, maltodextrin, modified sugars, and hydrogenated or interesterified oils, as well as so-called cosmetic additives designed to alter color, aroma, texture, or taste. Monteiro et al., 2019 — identifying Nova 4 foods

“Five or more ingredients” is a common clue, not a mathematical rule. Homemade soup can contain more than five ingredients without becoming ultra-processed; a drink may list only four and still have a typical Group 4 formulation. Likewise, the presence of any additive does not automatically establish Nova 4 or toxicological danger.

A presumed commercial purpose—convenience, longer shelf life, standardized flavor—is also insufficient when it cannot be observed. Accurate classification ideally requires the full ingredient list, sometimes process information, and explicit decision rules. When those data are missing, “undetermined group” is more honest than invented certainty.

• Learn how to identify ultra-processed foods in practice →

7. A product name, package, or marketing claim does not determine its group

Organic, plant-based, gluten-free, high-protein, and sports-oriented are not Nova groups. A protein bar may be Nova 4, but a packaged mix of plain nuts does not become Nova 4 simply because it carries a brand. Conversely, a product with artisanal-looking graphics may remain a complex industrial formulation.

A product can change groups after reformulation without changing its name. That is why a universal product list becomes outdated quickly. Collaborative databases and apps can offer a clue, but they depend on current, accurate ingredient information. Missing data should not be converted into a confident classification.

For an everyday decision, begin with three questions: Does a recognizable food remain the recipe’s foundation? Does the list contain fractions rarely found in a home kitchen? Are additives used mainly to reconstruct color, aroma, taste, or texture? If the answer remains unclear, first assess nutritional profile, serving size, and the product’s actual role in your diet.

8. Nova’s limitations matter, but they do not erase the entire tool

In an opinion signed in 2024 and published in 2025, ANSES noted that no consensus definition of ultra-processed food exists. It judged that Nova relies more on formulation than on the degree of processing, uses non-exhaustive lists of substances and additives, and contains no threshold or concept of hazard. The term “ultra-processed” can therefore create confusion. ANSES — opinion on the characterization and effects of so-called ultra-processed foods

This criticism does not mean Group 4 formulations do not exist or that every study using Nova is invalid. It calls for separating different dimensions: nutritional profile, food matrix, processes, additives, newly formed contaminants, packaging, and consumption conditions. Two Nova 4 products may differ along every one of those axes.

Usefulness also depends on context. At the population level, a broad category can track a shift from cooked meals toward ready-to-eat formulations. When deciding whether one yogurt suits one person, that breadth becomes a limitation. A tool can support monitoring without becoming an individual diagnostic test.

• Examine the evidence for and against Nova’s reliability →

9. Nova and Nutri-Score answer different questions, so their results can diverge

Nutri-Score mainly summarizes selected favorable and unfavorable components per 100 g or 100 mL. Nova focuses on formulation and processing. A product can therefore receive an A or B while belonging to Nova 4: its counted nutrients may produce a favorable profile even though the recipe contains industrial-formulation markers.

The reverse can also occur. Oil or cheese with a simple recipe may receive a less favorable nutritional grade on certain criteria without being Nova 4. The difference does not prove manipulation; the tools measure different dimensions.

When choosing between substitutable products, first consider the food’s role and expected amount. Then compare calories, protein, fiber, sugar, saturated fat, and salt according to your needs. Nova adds information about formulation; it should not replace overall nutrition, enjoyment, budget, or your ability to sustain the choice.

• Read Nutri-Score and Nova together without confusing them →
• Understand exactly what Nutri-Score summarizes →

10. Use Nova as a follow-up question, then examine your actual diet

For seven days, note the products that recur instead of judging one isolated purchase. For each, record its role in the meal, serving size, calories, and relevant macros, then check whether the recipe contains Nova 4 markers. This exercise highlights the main sources without turning every condiment into an investigation.

Next, choose one realistic replacement: water or an unsweetened drink instead of a daily soda, oatmeal instead of a highly formulated cereal dessert, or a simple batch-cooked meal instead of several ready-to-heat dishes. The replacement must preserve affordability, time, enjoyment, and enough protein to remain sustainable.

Nalko does not certify a product’s Nova group. The app uses the group available in its data, treats simple foods in its database as Nova 1, and assigns calculable calories from other entries without a group to “Nova undetermined.” Its internal average covers calories associated with Groups 1–4; it is neither an official Nova score nor a health verdict.

• Describe the recipe before assigning a group.
• Use “undetermined” when ingredient or process information is missing.
• Track frequency and quantity, not just the number of different products.
• First replace one frequent source with a feasible option.
• Re-evaluate after one to two weeks before adding a new rule.