Protein-to-calorie ratio: the foods that rank highest (and lowest)

Here's the full table, why the "high protein" foods at the bottom aren't actually a problem, and how to run this number yourself on anything in your fridge.

What the ratio actually measures (and the formula)

Protein-to-calorie ratio answers one narrow question: for every 100 calories a food costs you, how much protein do you get back? It's not a measure of overall nutrition, or whether a food is "good" or "bad" — it's a planning number for a specific job, the same way miles-per-gallon tells you about fuel efficiency and nothing about how comfortable the seats are.

The formula:

(protein grams ÷ calories) × 100 = protein per 100 calories

A food with 20g of protein and 200 calories scores 10. A food with 20g of protein and 100 calories scores 20 — twice as protein-efficient, even though both foods hand you the same 20g. That's the whole idea: total protein grams tell you what you got, but the ratio tells you what it cost you in calories to get it.

The full ranked table: best to worst

Figures below are rounded, commonly-cited values from USDA FoodData Central, the federal nutrient database.¹ Cooking method, cut, and brand shift the exact numbers a little — treat these as planning numbers, not lab-precision ones.

Food (cooked, per 100g)	Calories	Protein	Protein per 100 kcal
Shrimp	99	24g	24g
Canned tuna in water, drained	100	23g	23g
White fish (cod)	105	23g	22g
Turkey breast, skinless	135	30g	22g
Egg whites	52	11g	21g
Chicken breast, skinless	165	31g	19g
Nonfat Greek yogurt	59	10g	17g
90/10 ground beef	176	20g	11g
Tofu, firm	144	15g	10g
Edamame	121	12g	10g
Lentils	116	9g	8g
Bacon, pan-fried	541	37g	7g
Cheddar cheese	403	25g	6g
Almonds	579	21g	4g
Peanut butter	588	25g	4g

The gap between the top and bottom of this list is roughly sixfold — shrimp delivers six times more protein per calorie than peanut butter, despite peanut butter's reputation as a protein food. That gap is entirely about fat: fat carries 9 calories per gram against protein's 4, so a food doesn't need much fat before it starts diluting the ratio hard. Chasing a high ratio isn't just a numbers game, either — sports-nutrition research links adequate protein intake to better satiety and lean-mass retention while eating in a deficit, which is exactly the situation a protein-dense food is built for.²

Why some "high-protein" foods rank surprisingly low

Peanut butter, almonds, cheddar, and bacon all genuinely contain a lot of protein per serving — two tablespoons of peanut butter really does have around 7g. The problem for the ratio isn't the protein; it's everything riding alongside it. Almonds are roughly 50% fat by weight. Bacon renders most of its calories from rendered fat, not the meat itself. Cheddar is a dairy fat product that happens to also carry protein.

None of that makes these foods bad. It means the label "high protein" is doing less work than people assume, because almost anything with a meaningful amount of protein also has some fat riding along — and food marketing rarely mentions the denominator. A protein bar that boasts "20g of protein!" on the front can still be a 300-calorie bar, which puts it well below chicken breast on this exact measure, even though nobody printed that on the wrapper.

How to calculate the ratio for any food yourself

You don't need this article's table for foods that aren't on it — every packaged food already gives you both numbers on the label.

1. Find the grams of protein per serving.
2. Find the calories per serving.
3. Divide protein by calories, multiply by 100.

Try it on a granola bar: 4g protein, 190 calories → (4 ÷ 190) × 100 ≈ 2.1g per 100 kcal — lower than almost everything in the table above, which tells you plainly that a granola bar is a carb-and-fat snack that happens to contain a little protein, not a protein food. Compare that to a cup of low-fat cottage cheese: roughly 25g protein, 180 calories → (25 ÷ 180) × 100 ≈ 13.9g per 100 kcal — a real protein source by this measure.

A low ratio isn't a red flag

I almost left this section out of Bocco's own nutrition breakdown for exactly this reason — the first time I showed a beta tester a low protein-per-calorie number next to their peanut butter toast, they asked if the app was telling them to stop eating it. It wasn't, and I want to be honest about that limit here, because it's tempting to read a table like this as a hierarchy of good and bad foods. Olive oil scores 0 — it has no protein at all — and it's still one of the better fats you can cook with. Avocado scores low too, and it's doing real work for fiber and healthy fats that this specific number was never built to capture.

The ratio is useful for exactly one job: figuring out which foods let you hit a protein target without spending your whole calorie budget getting there. It's not a scoring system for your whole diet, and a food doesn't need to defend itself against it.

This is the exact math Bocco runs in the background every time it estimates a meal — protein grams and total calories together, not protein in isolation. If a portion estimate looks off, you correct it in a sentence ("that was a larger cut of chicken") and both numbers recalculate honestly, so the ratio you're seeing stays real instead of drifting.