Protein-to-Calorie Ratio in the USDA Database, Methodology Note

How GetFoodFacts computes protein-to-calorie ratios from USDA FoodData Central, what the metric tells you, and the limitations to keep in mind.

The protein-to-calorie ratio (often expressed as grams of protein per one hundred kcal) is one of the most useful single sorting heuristics in nutrition planning. It captures the central tradeoff in protein selection: maximising the amount of muscle-supporting amino acids you consume per unit of energy. The metric is straightforward to compute — divide grams of protein per one hundred g by kilocalories per one hundred g, then scale to a per-one-hundred-kcal denominator. Its simplicity is its strength.

Within the USDA FoodData Central dataset that powers GetFoodFacts (currently 2,013,644 entries, of which 1,999,950 sit in the Branded Foods sub-database), the foods that score highest on this metric tend to fall into a small number of categories: lean white fish (cod, tilapia, pollock, haddock), shellfish (shrimp, lobster, crab), pure egg whites, skinless poultry breast, and lean game meats (rabbit, venison, bison loin). These foods share a common nutritional profile — high lean-protein content paired with negligible fat per gram.

Why the metric matters in practice

Two scenarios make this metric especially useful. The first is calorie-controlled or hypocaloric eating — weight loss, body recomposition, athletic cuts — where every calorie spent on a non-protein macronutrient is a calorie taken from somewhere else in the budget. Foods at the head of the ranking deliver maximum protein for minimum energy expense, which research links to better satiety per kcal (Westerterp-Plantenga et al, 2009) and easier adherence over time. The second is budget protein meal planning — many of the highest-scoring foods (canned fish in water, eggs, frozen white fish) are also among the cheapest sources of complete animal protein per gram.

Methodology

GetFoodFacts pulls protein and energy values directly from FoodData Central's Standard Reference Legacy and Foundation Foods datasets. We exclude Branded Foods from the protein-per-calorie ranking because branded values are manufacturer-submitted from product labels, where rounding rules and partial nutrient panels reduce comparability. We also exclude entries with calorie values below ten kcal per one hundred g (water, herbs) so the denominator stays meaningful, and entries missing either protein or calorie measurements. The ranking updates automatically rather than being pre-baked into static text, so it stays current with the underlying data.

What the leading foods have in common

Foods that score highest on protein-per-calorie share four structural traits. First, they have minimal lipid content. Cooking oils, butter, mayonnaise, and processed snacks dominate the bottom of the ranking precisely because their energy density is built almost entirely from fat. Lean white fish and skinless poultry sit at the top because nature partitions their tissue heavily toward water and structural protein, with virtually no intramuscular fat. Second, the leaders have minimal carbohydrate content. This excludes legumes, grains, dairy with lactose, sweetened protein bars, and most plant proteins. Third, they have minimal moisture-replacing additives — no breading, no glazes, no sauces. Fourth, they are consistently sourced — chicken breast in one supermarket has nearly the same protein density as chicken breast in another, which simplifies meal planning and shopping-list construction.

These traits cluster naturally in lean fish (cod, haddock, halibut, pollock, sole, tilapia, perch), in shellfish (shrimp, scallops, lobster, crab, octopus, squid), in poultry breast cuts (chicken, turkey, duck breast, quail breast), and in selected lean game (rabbit, venison, ostrich, buffalo loin). Eggs are a reference protein in their whole form but their protein-per-calorie ratio is dragged down by the yolk lipids; isolating the white doubles the score. Lean cuts of beef and pork sit in the second tier — sirloin, tenderloin, eye-of-round — while marbled cuts and processed meats fall further down. Tofu, seitan, tempeh, edamame, and lentils represent the plant-protein strata; they trade off completeness of essential amino acids for accessibility, environmental footprint, and saturated-fat profile.

How GetFoodFacts ranks foods

Each food in the database has a USDA-assigned data_type (sr_legacy, foundation, branded, or survey), a category mapping (dairy, fish, vegetables, grains, etc.), a primary description, and a nutrient column for every measured nutrient. The protein-per-calorie ratio sorts foods by protein / calories with calories in the denominator, then multiplies by 100 to express as grams per 100 kcal. We exclude entries below 10 kcal per 100 g (water, herbs, spices) and entries with missing protein or calorie measurements. The result is a deterministic ranking that anyone with the same USDA dataset can reproduce.

The ranking is recomputed at server-side render time. There is no static cached ordering; refresh the page after a USDA dataset update and the rankings reflect the new values. Each food in the table links to its full FDA-format Nutrition Facts label, where you can inspect the per-serving values, micronutrient breakdown, allergen disclosures, and source citations. The methodology page documents the computation in full and links the source agency.

Limitations

Protein-per-calorie does not capture protein quality. The PDCAAS and DIAAS scoring systems both penalise plant proteins for incomplete amino acid profiles — particularly low lysine, methionine, threonine, or tryptophan — which a simple ratio does not reflect. Cooking method matters — the ranking uses raw values where USDA provides them, and deep-frying, breading, glazing, or sweetening can move a real-world serving far away from those numbers. Sodium content (canned tuna, smoked fish, cured meats, processed deli items), mercury content (large predatory fish like swordfish, king mackerel, tuna steaks, marlin), pesticide residues (conventional produce), preservative additives, and the presence of artificial colourants or sweeteners are not folded into the ranking. We list each food's source agency entry so you can review the full nutrient panel, ingredient list, and contamination caveats before adopting any of these as a regular dietary pillar.

The ranking is also a single snapshot in time. USDA periodically re-analyses foods when analytical methods improve or when reformulations change the underlying nutrient profile. A 2010 USDA value for a packaged food may not match the 2024 value if the manufacturer changed the recipe; we use the most recent value USDA provides at the time of our database build. Manufacturer-submitted Branded Foods values are not independently verified by USDA, so they reflect what was on the label, not what the food actually contains.

Related references

Sources