How We Grade Your Labels: The Science Behind the Score
Most food apps give you a score but never explain how they got it. Here's the methodology behind GradeMyLabel's Food Risk Score — transparent, evidence-based, and free from brand influence.
Most food apps give you a score. A number, a color, a letter grade. Few explain how they arrived at it. Even fewer let you verify their work.
We think that is a problem. If you are going to make decisions about what you feed yourself and your family based on a score, you deserve to understand how that score is generated. So here is our approach — transparent, evidence-based, and free from brand influence.
What We Analyze
GradeMyLabel’s Food Risk Score draws on multiple data sources, each chosen for scientific credibility and regulatory authority.
Regulatory databases from 6 countries. We track the approval status of every ingredient across the United States, European Union, China, Japan, India, and South Korea. These six regions represent the world’s major food regulatory frameworks, each with its own scientific review process and risk tolerance. When regulators disagree, that disagreement itself is meaningful data.
EFSA safety assessments. The European Food Safety Authority publishes detailed toxicological evaluations including Acceptable Daily Intake values and genotoxicity assessments. We incorporate these directly into our database.
Published peer-reviewed research. Our database references published studies from journals including Food and Chemical Toxicology, The Lancet, and other indexed publications. We track study conclusions, journal quality, and whether findings have been replicated.
Our database currently covers 582 ingredients across 48 categories — from preservatives and colorants to emulsifiers, sweeteners, and sunscreen filters.
The Three Pillars of Our Score
The Food Risk Score evaluates each ingredient through three independent lenses, then combines them into a single score for the overall product. Here is how each pillar works conceptually.
Pillar 1: Research Risk Level
Every ingredient in our database is classified based on the weight of published scientific evidence: high risk, moderate risk, low risk, or under review. This classification is not based on a single study or a headline. It reflects the overall body of evidence — multiple studies, from multiple research groups, published in peer-reviewed journals.
An ingredient classified as “high risk” has consistent evidence of health concerns across multiple independent studies. “Moderate risk” means there is credible evidence of concern, but the picture is incomplete or the effects are dose-dependent. “Low risk” means the weight of evidence supports safety at typical consumption levels. “Under review” means active scientific evaluation is ongoing and the evidence is not yet conclusive.
Pillar 2: Regulatory Consensus
A single country banning an ingredient might reflect that country’s particular risk tolerance or political environment. But when multiple countries independently evaluate the same evidence and reach the same conclusion, that consensus carries real weight.
We cross-reference regulatory decisions across all six countries in our database. An ingredient that is banned or restricted in several countries receives a higher risk signal than one that is restricted by a single outlier. The pattern of global regulatory agreement — or disagreement — is itself a powerful data point.
Importantly, your score is personalized. Your country’s regulatory stance is weighted more heavily in your individual score than other countries’ positions. If you live in the EU and an ingredient is approved there but banned in Japan, that context is reflected differently than if you live in Japan. Your local regulatory environment matters most to your daily choices.
Pillar 3: Known Health Risks
Beyond the broad risk classification, we evaluate the specific health effects associated with each ingredient — what they are, how severe they are, and how strong the evidence is. An ingredient linked to mild gastrointestinal sensitivity in a single study is treated differently from one linked to carcinogenicity across multiple replicated studies.
The severity of documented effects and the quality of evidence behind them both factor into the score. Not all risks are equal, and our scoring reflects that reality.
What We Do NOT Do
What we exclude from our methodology is just as important as what we include. These are deliberate choices, not oversights.
We do not score nutrition. GradeMyLabel measures ingredient safety — whether the additives, preservatives, colorants, and processing aids in a product pose health risks. Nutritional quality is a separate question, and a good one, but it is not our question. For barcode scans where nutritional data is available, we display the Nutri-Score alongside our Food Risk Score so you get both perspectives without us conflating them.
We do not give organic products a bonus. “Organic” is a production method certification, not a safety assessment. An organic product can contain ingredients that are genuinely concerning. A conventional product can be perfectly safe. Our score is based on what is actually in the product, not how it was produced.
We do not treat all additives as equally bad. This is perhaps our biggest philosophical difference from some competitors. Citric acid and Red Dye 3 are both “additives.” One is a natural component of citrus fruit that your body produces on its own. The other was banned by the FDA in 2025 after decades of evidence linking it to cancer in animals. Treating them as equivalent because they are both “additives” is not scientific — it is fearmongering.
We do not accept money from brands. No brand can pay to improve their score. No partnership affects our risk assessments. Our scoring methodology is applied uniformly to every product we evaluate.
Household Profiles
Not all ingredients are equally risky for all people. A preservative that is perfectly safe for a healthy adult may pose different risks for an infant whose liver enzymes are still developing, or for a pregnant person whose metabolic pathways are altered.
GradeMyLabel supports household profiles with scoring adjustments for three age groups — infant, child, and adult — plus a pregnancy modifier. When you scan a product with a child’s profile active, the score reflects age-appropriate risk assessments drawn from regulatory guidelines like EU Regulation 1333/2008, which sets specific additive restrictions for foods marketed to young children.
We also support pet profiles for dogs and cats. Many common food ingredients — xylitol, theobromine, certain allium compounds — are safe for humans but toxic to pets. Pet profiles flag species-specific toxins when you scan products, whether you are checking pet food or considering sharing human food with your animal.
EFSA Integration
For 76 ingredients, we display the Acceptable Daily Intake — the amount a person can safely consume every day for a lifetime, as determined by EFSA’s scientific panel. This is real toxicological data, with the built-in 100-fold safety margin that EFSA applies.
We also show genotoxicity assessments for 105 ingredients — whether EFSA has found evidence that a substance can damage DNA. This is the same data that European regulators use when deciding whether to ban a substance, and we make it available to you in plain language at the point of purchase.
To our knowledge, no other consumer food safety app surfaces ADI and genotoxicity data directly. Most apps stop at “good” or “bad.” We show you the numbers behind the judgment.
Transparency and Community
We believe scoring methodology should be auditable, and our database should grow with community input.
User corrections. If you think an ingredient is scored incorrectly, you can flag it directly from the scan result screen. Every correction is reviewed by our team against the published evidence before any change is made.
AI-powered research pipeline. When our system encounters an ingredient it does not recognize, it enters our research queue. Our admin team uses an AI-assisted research pipeline that pulls from regulatory databases and published literature to generate a comprehensive safety profile, which is then reviewed by a human before publication.
Regular database updates. New ingredient data, study references, and regulatory changes are delivered to your device automatically via over-the-air patches. You always have the latest data without needing to update the app.
Our database grows every week. We started with common additives and now cover 582 ingredients. Every one of them has been individually researched, classified, and cross-referenced against six countries’ regulatory positions.
The Bottom Line
A score is only as good as the methodology behind it. Ours is built on regulatory data from six countries, EFSA toxicological assessments, published peer-reviewed research, and a scoring model that considers dose, evidence quality, and regulatory consensus — not just whether an ingredient sounds “chemical.”
We are not perfect. No scoring system can capture every nuance of food safety science. But we are transparent about our approach, rigorous about our data sources, and committed to showing you the evidence rather than just the conclusion.
Download GradeMyLabel and see the science behind every product you scan. Real data, real sources, no brand sponsorships — just the information you need to make informed choices.