Complete vs. Incomplete Proteins: What the Data Actually Shows

A protein counts as "complete" if it supplies all nine essential amino acids your body cannot make itself; an "incomplete" protein falls short on at least one. Measured values in this database show soy nearly matches egg's lysine level gram-for-gram of protein, but delivers about a third of egg's methionine — the gap behind the label.

What actually makes a protein "complete" instead of "incomplete"?

The split comes down to nine specific amino acids — histidine, isoleucine, leucine, lysine, methionine, phenylalanine, threonine, tryptophan, and valine — that the human body cannot synthesize on its own and must get from food. As MedlinePlus puts it plainly: "Proteins from meat and other animal products are complete proteins. This means they supply all of the amino acids the body can't make on its own. Most plant proteins are incomplete." Harvard T.H. Chan School of Public Health, The Nutrition Source defines a complete protein the same way — one containing all of the essential amino acids needed to build new protein in the body — and confirms the same nine-amino-acid list.

That framework explains the category, but it doesn't tell you how close an "incomplete" food actually comes, or which amino acid is really the bottleneck. For that, you need the measured values, not just the label.

Which foods in this database are actually complete proteins?

Per Harvard's Nutrition Source, animal-based foods are reliably complete, and a short list of plant foods — soy and quinoa among them — also qualify as complete on their own, without needing to be paired with anything else. Checking this database's own values for protein confirms why soy keeps showing up on that short list: a 100-gram serving of defatted soy flour carries 51.1 g of protein, versus 12.4 g in a 100-gram serving of whole egg — soy flour is simply a far more concentrated protein food, which is part of why raw gram-for-gram comparisons of specific amino acids can mislead unless you normalize for how much protein is actually in the food.

Does "incomplete" mean a food is a lesser protein source?

This is where normalizing the data changes the picture. Looking at lysine and methionine content relative to each food's own protein content (grams of amino acid per gram of protein, calculated from this database's measured amounts) tells a sharper story than the raw per-100g numbers:

• Whole egg: roughly 67 mg of lysine and 34 mg of methionine per gram of protein.
• Defatted soy flour: roughly 60 mg of lysine per gram of protein — close to egg — but only about 12 mg of methionine per gram of protein, roughly a third of egg's level.
• Raw almonds: roughly 29 mg of lysine and 7 mg of methionine per gram of protein — lower than egg on both counts.
• Raw walnuts: roughly 26 mg of lysine and 13 mg of methionine per gram of protein.

That pattern matches the classic legume profile: soy is unusually lysine-rich for a plant food (which is exactly why it qualifies as "complete"), but like most legumes it's still relatively short on sulfur-containing amino acids such as methionine compared with egg. Tree nuts don't all behave the same way, either — almonds run lower than soy on both amino acids relative to their own protein content, but walnuts land closer to soy's methionine density than almonds do, despite both being "just nuts" in casual conversation. That spread is the real point: "incomplete" isn't one flat category sitting below "complete," it's a gradient, and two foods with the same informal label (nut, legume, grain) can sit in noticeably different places on it. None of this makes almonds, walnuts, or soy "worse" foods — it just means the amino acid gap an "incomplete" label is flagging is real and specific, not a vague quality knock.

Are all "incomplete" proteins incomplete in the same way?

Not quite, and the amino acid itself matters as much as the category. MedlinePlus separates amino acids into three groups: essential (must come from food), nonessential (the body can synthesize them), and conditional (normally nonessential, but required from food during illness or physical stress). The "complete vs. incomplete" framework only concerns the essential group, which is why a food can be extremely protein-rich by total grams and still be labeled incomplete — the issue was never quantity, it was which specific essential amino acids were present and in what ratio. Walnuts and almonds are a useful pair for seeing this: walnuts have less total protein per 100 g than almonds (14.6 g versus 21.5 g), yet walnuts edge out almonds on methionine density per gram of protein — total protein content and amino acid completeness are two different measurements, and a food can rank differently on each one.

Do dry-ingredient numbers overstate how much protein you're actually getting?

A comparison worth flagging before you start adding up grams from a label: this database lists dry lentils at 23.6 g of protein per 100 g, but once boiled, that same lentil measures only 9.0 g of protein per 100 g — the drop is water absorbed during cooking, not protein lost. Dry chickpeas show the same pattern: 21.3 g per 100 g dry versus 8.9 g per 100 g cooked. Cooked quinoa measures even lower, at 4.4 g of protein per 100 g, because it's a grain that's mostly water and starch once prepared. If you're comparing a dry-ingredient figure on one food against a cooked-weight figure on another, you're not actually comparing protein density — you're comparing hydration.

Do you really have to combine proteins in the same meal?

This is the single most persistent myth attached to incomplete proteins, and it's been corrected for decades. The old "protein combining" advice held that a vegetarian eating rice needed to eat beans in the same sitting to get a usable complete protein. Current guidance from Harvard's Nutrition Source states that people who avoid animal foods "can eat a variety of protein-containing plant foods each day in order to get all the amino acids needed" — there's no same-meal requirement in that guidance. MedlinePlus makes the same point about amino acid timing more broadly, noting that essential amino acids don't need to come from every single meal because "the balance over the whole day is more important." If you eat lentils at lunch and rice at dinner, your body still has access to both amino acid pools when it builds new protein — it doesn't require them to arrive on the same plate.

How much protein do you actually need, and does the source matter?

The National Academy of Medicine's figure, as cited by Harvard's Nutrition Source, is 0.8 grams of protein per kilogram of body weight per day, with an acceptable range of 10% to 35% of total daily calories. On the question of whether source matters beyond amino acid completeness, the same source notes that "'Pure' protein, whether derived from plant or animal foods, probably has similar effects on health, although the mix of amino acids can have health implications" — which is exactly the kind of gap this guide has been measuring directly rather than describing vaguely.

For a side-by-side look at how real animal and plant foods in this database stack up on protein content once you account for how they're actually prepared and eaten, see our companion comparison of plant-based and animal protein sources.

So does "incomplete" protein mean you need to change what you eat?

"Complete" and "incomplete" are not a verdict on food quality — they're a description of one specific nutritional dimension: whether a food, eaten alone, covers all nine essential amino acids. Soy earns its complete-protein status on real data, not marketing. Most legumes fall short mainly on methionine, not across the board. And the complete-protein framework has never required same-meal pairing to work — eating a variety of protein sources across your day is what the current evidence actually supports.

---