Protein Sources Ranked by Bioavailability, Cost, and Convenience

The Numbers in Every Other Ranking Are Mostly Fake

Search for a protein source ranking and you will find the same table everywhere. Whey isolate 1.09. Chicken breast 1.08. Whole egg 1.13. Beef 1.11. Peanuts 0.43. Almonds 0.40. The numbers are quoted to two decimal places, which reads like precision, and they show up in enough places that they feel settled.

We went looking for where they came from. The chicken number is the instructive one.

The DIAAS of 1.08 for chicken breast appears in a 2019 Sports Medicine supplement table, footnoted "measured in pigs." That footnote is wrong. The cited source is Ertl and colleagues, published in 2016 in Die Bodenkultur, an Austrian agronomy journal, in a paper about the land occupancy of livestock products. Its Table 5 lists 108.2 for "chicken meat and skin," which is already a different food from a skinless breast. And the paper's own methods section says the quiet part out loud: because ileal amino acid digestibility values for these products did not exist, the authors substituted a single crude protein digestibility figure from a 1991 FAO report, 94 percent for meat, 95 percent for milk, 97 percent for eggs, and applied it uniformly to every amino acid.

That is the PDCAAS method. Applying one blunt whole-protein digestibility coefficient to all nine indispensable amino acids is the specific flaw that DIAAS was designed to fix. The number has been laundered through a sports nutrition review into a hundred fitness blogs, and it now sits in tables labelled "measured." The same source is the origin of the circulating beef 1.09, pork 1.14, and whole egg 1.16 figures.

Almonds are worse. There is no DIAAS for almonds anywhere in the literature. The definitive study on almond protein quality reports a PDCAAS of 44.3 to 47.8 with lysine limiting, and its own discussion says DIAAS "may prove challenging to implement," meaning the authors did not calculate one. Every "almond DIAAS 45" online is that PDCAAS with a new label.

None of this means chicken is a bad protein. It obviously is not. It means the confident-looking decimal in the table you read was decoration. This article uses only values that came out of a real digestibility measurement, and it tells you which model produced each one.

What DIAAS Actually Measures

The Digestible Indispensable Amino Acid Score was recommended by the FAO in 2013 to replace PDCAAS. Both scores answer the same question: how well does the amino acid profile of this food match what a human needs, after accounting for what actually gets absorbed. They differ in two ways that turn out to matter.

First, PDCAAS uses faecal crude protein digestibility. Faecal measurement is contaminated by gut bacteria, which ferment and modify amino acids in the large intestine after the point where you can absorb them. That systematically overstates how much protein you got. DIAAS uses ileal digestibility, sampled at the end of the small intestine, measured per amino acid rather than for the protein as a whole.

Second, PDCAAS is truncated at 1.00. Any protein scoring above the reference pattern is recorded as 1.00, which erases the difference between a food that just clears the bar and one that clears it by 40 percent, and it makes it impossible to calculate how a high-scoring food compensates for a low-scoring one in a mixed meal. DIAAS is not truncated.

Three details determine whether a DIAAS number means anything:

1. Which reference pattern. FAO publishes three: birth to six months, six months to three years, and older children through adults. The same food scores very differently across them. Whey protein isolate is 67 on the infant pattern and 125 on the adult pattern. Any DIAAS quoted without its pattern is unusable, and most online tables do not state one.
2. Which model. Ileal-cannulated pigs are the accepted proxy for humans. Growing rats are a weaker proxy and often produce much lower values. In vitro INFOGEST digestion is newer and cheaper. Some published "DIAAS" values are computed entirely from animal-feed composition tables with no digestibility measurement at all.
3. Whether the food was cooked the way you eat it. Cooked ground beef and raw ground beef differ by 22 points. Tofu and soy milk differ by 20.

FAO's classification bands: 100 or above is an "excellent" or "high" quality protein source, 75 to 99 is "good," and below 75 gets no quality claim. Those cutoffs were chosen by committee rather than derived from outcome data, which the FAO report itself concedes.

The Measured DIAAS Table

Everything below came from a published digestibility trial. The reference pattern is the older child, adolescent and adult pattern in every row, so the values are comparable to each other. Model and source are stated because they are not interchangeable.

Three honest caveats. The bean and hemp values from Nosworthy and colleagues used true faecal nitrogen digestibility, applying one coefficient across amino acids, which the papers state openly and FAO permits as an interim method. Treat those rows as DIAAS-like rather than equivalent to the pig-derived rows above them.

Cereal grains appear here only as whole-grain wheat at 54. A separate rat study of nine cooked cereals reported far lower values, including whole wheat at 20, oats at 43, brown rice at 42, and millets in single digits. Those numbers conflict badly with the pig literature, and the paper's own footnotes contradict its methods on which reference pattern was used. Grains are clearly the weakest protein category. The precise depth of the hole is unresolved.

Potato protein is the outlier worth watching, scoring around 100 on the young child pattern in a meta-analysis of pig digestibility data and being the only plant isolate in a widely cited amino acid analysis that met WHO requirements for every essential amino acid. That meta-analysis was written by four employees of a potato protein company. The independent in vitro number is 87 with valine limiting.

Limiting Amino Acids and Why Single-Food Rankings Mislead

DIAAS scores a food on its worst amino acid. A protein could be generous in eight of the nine indispensable amino acids and score 54 because it is short on lysine. Whole-grain wheat is exactly that food.

Which means a ranked list of single foods answers a question almost nobody has. You do not eat wheat alone. You eat it with dairy, or eggs, or beans, or meat. And the moment a second protein enters the meal, the limiting amino acid of the first one may stop limiting anything.

The pattern is consistent enough to be useful:

Grains run out of lysine. Legumes run out of methionine and cysteine. Combine them and the mixture scores higher than either component. That is the entire mechanism behind rice and beans, hummus and pita, dal and roti, and every other grain-legume staple pairing that emerged independently on four continents.

It has limits. A computational optimisation of pea and rice protein mixtures found the best achievable DIAAS was 84, at 41 percent rice. Better than rice alone by a wide margin, still short of 100. The frequently marketed claim that a rice and pea blend hits 100 has no measured trial behind it and traces to a corporate press release about pea protein on its own.

Cost Per 30 Grams of Protein

This is the axis that changes your behaviour, because you pay it every week for the rest of your life.

The four rows marked BLS use official June 2026 US city average retail prices from the Bureau of Labor Statistics average price series. The rest are representative national retail prices as of September 2026 and will vary by store and region. Protein content per food comes from USDA composition data. The arithmetic is shown so you can redo any row with your own receipt.

The spread is more than sixfold from top to bottom, and the ordering is not what most lifters assume.

Eggs are the best value in high quality protein, and it is not close. At 85 cents per 30 grams they beat whey powder, cost a third of what ground beef costs, and carry the highest measured DIAAS of any whole food in the table. Egg prices fell 2.3 percent from May to June 2026 and USDA projects them down more than 30 percent across the year, while beef went the other direction and is up 72 percent against January 2021. If you set your protein habits during the 2022 egg spike, the arithmetic has changed underneath you.

Ground beef has quietly become a luxury protein. At $2.62 per 30 grams it costs three times what eggs cost for a cooked DIAAS of 99. Still excellent food, with iron, zinc, B12, and creatine attached. Still a poor way to buy protein per dollar in 2026.

Whey powder is competitive on price, which surprises people who file it under premium products. At a dollar per 30 grams from a warehouse club it undercuts chicken breast. The powder buyer's guide covers what separates a good tub from an expensive one.

Legumes are absurdly cheap and the DIAAS penalty is real but small in context. Lentils deliver protein at 42 cents per 30 grams, one fifth the price of Greek yogurt. Score them at 68 instead of 130 and they still win on cost per unit of usable protein by a wide margin. The genuine cost is that the protein arrives attached to 100-plus grams of carbohydrate, a real constraint during a cut and a feature during a bulk.

What the Muscle Data Says About Plant vs Animal

Here is where the ranking gets complicated, because the outcome data does not track the quality scores as tightly as the scores imply it should.

The theoretical case against plant protein is well documented. A 2018 analysis of ten commercial plant isolates found leucine ranging from 5.1 percent of protein in hemp to 8.3 percent in potato, against 9.0 percent in milk and 11.0 percent in whey. Total essential amino acid content averaged 26 percent in plant isolates against 37 percent in animal isolates. Wheat, corn, oat, brown rice, hemp, and lupin all fell below WHO lysine requirements. Oat, lupin, and wheat missed on both lysine and methionine. Only potato protein met the requirement for every essential amino acid.

So the prediction is that plant protein should produce a weaker anabolic signal. Then the van Loon group at Maastricht went and tested it with stable isotope infusions and muscle biopsies, one plant source at a time, at a 30 gram dose in healthy young men.

Five trials, five nulls. In the blend study, milk delivered nearly twice the essential amino acid exposure over five hours (151 vs 79 mmol per 300 min per litre) and the plant blend still produced a numerically higher synthesis rate. Wheat, the protein that anchored the original "plant protein is inferior" argument, matched milk.

Longer trials point the same way. In 38 young men trained twice weekly for 12 weeks with protein equalised at 1.6 g/kg using soy isolate or whey, leg lean mass rose 1.2 kg in both groups, with no between-group difference on any measure including fibre cross-sectional area and leg press 1RM. A separate 10-week trial at roughly 2 g/kg with five training sessions per week found lean mass gains of 2.6 kg on an omnivorous diet and 3.1 kg on an exclusively non-animal diet, again with no significant difference.

The counterweight, which most articles on this topic omit, is that the largest meta-analysis to date still finds a small residual advantage for animal protein. Pooling 43 randomised trials, plant protein produced slightly less muscle mass (standardised mean difference -0.20), with no difference in strength or physical performance. The important detail is where that effect lives: across 17 trials there was no difference at all between soy and milk protein, and the entire animal-protein advantage came from non-soy plant proteins (SMD -0.58 across five trials) and whole plant-based diets (-0.51 across seven trials).

Where Protein Quality Genuinely Bites

The evidence above does not make source irrelevant. What it shows is that source stops mattering under a specific set of conditions, and most lifters reading this happen to meet them. When those conditions break, quality re-enters the picture.

• Low total intake. Every null result above came from a trial feeding 1.6 to 2.0 g/kg. At 0.8 g/kg the margin for a poor amino acid profile is much thinner, because there is no surplus of anything to cover a shortfall. Quality matters most exactly when quantity is scarce.
• Small doses. The plant-equals-milk findings used 30 gram servings. A 15 gram serving of a low-leucine plant isolate is a different proposition, because you may fail to reach the leucine threshold that triggers the synthesis response.
• Single-source plant diets. The one place the meta-analysis found a real deficit was non-soy plant protein and whole plant-based diets. A diet where the protein is overwhelmingly rice, oats, and wheat has an unresolved lysine problem no amount of total volume elegantly fixes.
• Aggressive cuts. Calories are scarce, protein needs go up rather than down, and every gram has to work. A cross-sectional comparison of vegetarian and omnivorous endurance athletes found DIAAS 11 percent higher and available protein 43 percent higher in the omnivores, with vegetarians estimated to need 10 to 22 extra grams a day to match. That study cannot prove causation, but the direction is consistent.
• Older lifters. Anabolic resistance raises the per-meal protein dose needed to trigger synthesis, which makes leucine density more relevant rather than less.

And one honest limitation: no trial has ever directly tested the interaction. Nobody has randomised subjects to plant or animal protein at both a low and a high total intake and shown the source effect shrinking as intake rises. That claim is an inference from converging evidence, not a measured result. The 1.62 g/kg plateau from Morton's 2018 meta-regression, which gets cited constantly in this context, is a threshold for protein quantity and says nothing about source.

The Leucine Question

Leucine gets treated as the master switch, and it does have a real mechanistic role in triggering the muscle protein synthesis response. The practical implication is usually overstated.

The measured leucine content of common isolates, as a percentage of total protein: whey 11.0, milk 9.0, potato 8.3, casein 8.0, egg 7.0, human skeletal muscle 7.6. Plant isolates averaged 7.1 percent against 8.8 for animal. Hemp came in at 5.1 percent and lupin 5.2, both below the WHO requirement. Corn was the outlier at 13.5 percent, higher than whey.

The blunt way to use that data is a conversion: to match the 2.7 grams of leucine in 25 grams of whey you would need roughly 20 g of corn protein, 33 g of potato, 37 g of brown rice, 38 g of pea, 40 g of soy, 45 g of wheat, 47 g of oat, 52 g of lupin, or 54 g of hemp. Useful as a sanity check on the outliers, and not a prescription, because the pea study delivered 30 grams, undershot milk on leucine by a wide margin, and produced an identical synthesis rate.

Leucine behaves like a threshold rather than a dial. Clear it and the response fires. The dose that clears it is comfortably reached by a normal 30 to 40 gram serving of nearly any protein except hemp or lupin, which is also why isolated branched-chain amino acid supplements do nothing useful. Full teardown in the BCAA article.

Combining Sources Without Doing Arithmetic

The 1970s version of complementary proteins told people to combine specific plant foods within a single meal. That has been abandoned. Your body maintains a free amino acid pool and draws on protein eaten across the day, so a rice lunch and a lentil dinner combine fine.

What still holds is that the combining has to happen at all. Four rules that cover it:

1. Grains plus legumes, at whatever interval. Lysine from the legume, sulfur amino acids from the grain. This is the single highest-yield move available to a plant-based lifter.
2. One high-scoring source per day fixes a lot. Because DIAAS is untruncated, a food scoring 135 carries surplus amino acids that offset deficits elsewhere. Two eggs, a glass of milk, or a scoop of whey does more repair work across a mixed diet than the calories suggest.
3. Soy and potato protein are the plant sources that need no partner. Soy isolate at 98 with a near-complete profile behaves like an animal protein in every outcome trial. Potato protein is the only plant isolate meeting WHO requirements for all essential amino acids.
4. Do not build a protein intake out of grains and nuts. Both are limited by lysine, so combining them fixes nothing. This is the actual failure mode in poorly constructed vegan diets, more than any total-intake problem.

Meal builds that solve the problem without anyone doing arithmetic: lentil soup with bread, rice and beans, oatmeal made with milk, peanut butter on wholegrain toast with a glass of milk, tofu stir fry with rice, chickpea curry with roti. Every one is a traditional dish that has existed for centuries, which is not a coincidence.

The Convenience Axis Nobody Scores

DIAAS has a number. Cost has a number. Convenience has no score, and it decides more diets than the other two combined. A protein source that takes 25 minutes to cook, keeps for three days, and has to be portioned in advance is competing against one that requires unscrewing a lid. Over a year, the compliance gap between those two dwarfs any DIAAS gap between them.

The pattern worth acting on: the cheapest sources are the least convenient, and the gap closes almost entirely with batch cooking. A pound of dried lentils is 90 minutes of unattended simmering once a week for 113 grams of protein at 42 cents per 30. Canned beans skip the cooking and roughly triple the cost, which is still cheap. Boiling a dozen eggs on Sunday converts the highest quality protein in the table into grab-and-go food. The systems for this are in the meal prep guide.

The Actual Rankings

Combining all three axes, weighted for a lifter trying to hit 1.6 to 2.2 g/kg without going broke or quitting.

Tier 1: anchor your diet on these

Eggs. Highest measured whole-food DIAAS at 135 to 137, second-cheapest high quality source at 85 cents per 30 grams, portable once boiled, and the most satiating food per calorie in the table. Prices fell through 2026 while beef rose.

Dairy: milk, Greek yogurt, cottage cheese, whey. DIAAS 123 to 141. Whole milk at $1.04 per 30 grams is one of the best values available and whey from a warehouse club matches it. Yogurt and cottage cheese cost more but need zero preparation and are extremely satiating. Casein digests slowly, which matters much less than the internet believes, as covered in the protein timing article.

Pork loin and chicken breast. Both above 100 in every measurement of comparable meats, both around $1.10 to $1.25 per 30 grams, both requiring real cooking. Pork loin is consistently underpriced relative to chicken and nobody buys it.

Tier 2: excellent with a caveat

Canned fish. Tuna and salmon are above 100, shelf-stable for years, need no preparation, and cost $1.43 to $1.95 per 30 grams. Salmon brings omega-3s. Tuna brings a mercury ceiling that caps sensible intake at a few cans a week.

Soy: tofu, tempeh, edamame, soy isolate. DIAAS 97 to 98, no measurable difference from whey in 17 pooled randomised trials, and tofu at $1.32 per 30 grams is cheaper than Greek yogurt. The best plant protein by a wide margin.

Beef. Nutritionally excellent, up 72 percent in price since January 2021, and $2.62 per 30 grams makes it the most expensive protein most lifters buy regularly.

Tier 3: cheap volume, needs a partner

Lentils, beans, chickpeas. DIAAS 53 to 88, cost 42 to 55 cents per 30 grams, enormous fibre payload, high satiety. The carbohydrate load is the real constraint. Pair with a grain or a Tier 1 source across the day and the amino acid problem disappears.

Pea and rice protein isolates. DIAAS 73 and lower individually, around $1.03 per 30 grams, and a blend tops out around 84 rather than the 100 the marketing claims. Usable if dairy is off the table. If it is not, whey costs the same and scores 50 points higher.

Tier 4: not a protein source

Grains, nuts, seeds, and collagen. Wheat at 54, oats and rice lower, peanuts around 43, hemp 43 to 45. These are foods with protein in them rather than protein sources, and counting them toward a target inflates your effective intake. Collagen is the sharpest case, containing zero tryptophan and negligible leucine, which makes it useless for muscle regardless of the label. Details in the collagen guide.

What Cooking Does to Protein Quality

Heat changes protein digestibility in both directions, and the effect is bigger than most people expect.

The cleanest data comes from beef fed to ileal-cannulated pigs. Raw ground beef scored 121 on the adult pattern and cooked ground beef scored 99. A ribeye roast scored 111 at 56 C, peaked at 130 at 64 C, and fell to 107 at 72 C. The authors concluded that fermenting, curing, drying, and cooking to minimally safe internal temperatures do not reduce protein quality, but that grinding meat or cooking with direct heat can, especially with overcooking. Grinding increases surface area, which increases exposure to the heat-driven Maillard reactions that lock up lysine.

Eggs go the other way, which is a useful reminder that "cooking destroys nutrients" is too simple. Boiled whole eggs scored 135, fried 135, and scrambled 137. Raw egg white contains avidin and is markedly less digestible than cooked egg. Cook your eggs.

For plants, heat generally helps by destroying the trypsin inhibitors and lectins that block digestion in raw legumes. Processing method matters within that: black beans scored 65 extruded, 63 cooked, and 53 baked. Tofu scores about 97 while soy milk scores about 117, a gap driven by lower sulfur amino acid digestibility in the coagulated product.

Common Mistakes

Quoting DIAAS values without their reference pattern. Whey isolate is 67 on the infant pattern and 125 on the adult pattern. A table that mixes patterns produces nonsense rankings, and most online tables do not state which one they used because whoever built them did not know it was a variable.

Treating a protein ranking as a shopping list. The ranking scores single foods eaten alone, which nobody does. Once two protein sources land in the same day, the limiting amino acid of the weaker one often stops limiting anything.

Optimising quality while missing on quantity. Switching from chicken to a marginally higher-DIAAS source changes almost nothing. Going from 1.0 to 1.6 g/kg changes a great deal. Get the total right first, using the daily intake guide, then think about sources.

Counting grains and nuts toward the protein target. Three slices of bread and two tablespoons of peanut butter looks like 20 grams on a tracking app. On usable amino acids it behaves like considerably less, and both foods are limited by the same amino acid so they do not complement each other.

Buying expensive protein out of habit rather than arithmetic. Ground beef rose 72 percent since 2021 while eggs fell through 2026. If your protein rotation was set three years ago and never revisited, you are probably overspending on at least two items.

Assuming the plant-versus-animal question is settled in either direction. Five acute trials found no difference at 30 grams and two feeding trials found no difference at 1.6 to 2.0 g/kg, while the largest meta-analysis still found a small advantage for animal protein on muscle mass, concentrated in non-soy plant sources. Anyone telling you this is closed is selling something.

Frequently Asked Questions

What is the highest quality protein source?

By measured DIAAS on the adult reference pattern, milk protein concentrate scores 141, whey protein concentrate 133, whey protein isolate 125, skimmed milk powder 123, and cooked whole eggs land between 135 and 137 depending on preparation. Beef ribeye roasted to 64 C scores 130. Those all came out of actual digestibility trials in cannulated pigs. If the question is which single food gives the most usable amino acids per gram of protein, dairy concentrates and eggs win. If the question is which food should anchor your diet, quality is one input of three, and cost and convenience usually decide it.

Is the DIAAS score for chicken breast really 1.08?

No, and there is no measured DIAAS for chicken at all. The 1.08 traces through a 2019 Sports Medicine table to a 2016 paper in Die Bodenkultur, an Austrian agronomy journal, which listed 108.2 for chicken meat and skin. That paper states in its own methods that ileal digestibility values were unavailable, so it substituted a single 1991 faecal crude protein coefficient and applied it uniformly to every amino acid, which is exactly the calculation DIAAS replaced. The only chicken figure with a real measurement behind it is an in vitro INFOGEST result of 113.3 on the young child pattern. Chicken is clearly a high quality protein. The specific number is not evidence of it.

How much more plant protein do I need to eat to match animal protein?

Less than the DIAAS gap implies, and it depends on which plant food. A cross-sectional comparison of 38 omnivorous and 22 vegetarian endurance athletes found DIAAS 11 percent higher and available protein 43 percent higher in the omnivores, estimating the vegetarians needed 10 to 22 extra grams per day. That study is observational. On the other side, controlled feeding at 1.6 to 2.0 g/kg found no difference in lean mass or strength over 10 to 12 weeks. A practical rule is to add 10 to 20 percent to your target if the great majority of your protein comes from grains, legumes, and nuts rather than soy, potato protein, or dairy.

What is the cheapest way to hit a high protein intake?

At June 2026 US city average prices, dried lentils and beans deliver 30 grams of protein for 42 to 55 cents, far cheaper than anything else, though the protein arrives with a large carbohydrate and fibre load and a DIAAS in the 60s to 80s. Among high quality sources, eggs are the standout at about 85 cents per 30 grams at $2.14 a dozen. Whey from a warehouse club runs close to a dollar, whole milk $1.04, boneless chicken breast $1.23. Ground beef is the expensive one at $2.62 after a 72 percent price increase since January 2021.

Does plant protein build less muscle than whey?

For soy specifically, no. A 2018 meta-analysis of nine trials in 266 participants found no difference between soy and whey or other animal proteins for strength or lean mass, and a 2025 meta-analysis of 43 randomised trials found no pooled difference between soy and milk protein across 17 trials. For plant sources other than soy, a small advantage for animal protein does appear, with a standardised mean difference around -0.58 across five trials, showing up for muscle mass rather than strength. At a single 30 gram dose, controlled infusion studies found corn, pea, potato, wheat, and a plant blend all produced synthesis rates statistically indistinguishable from milk.

Does cooking reduce protein quality?

It can. In ileal-cannulated pigs, raw ground beef scored 121 on the adult pattern while cooked ground beef scored 99. Ribeye roasted to 56 C scored 111, to 64 C scored 130, and to 72 C scored 107. The authors concluded that fermenting, curing, drying, and cooking to minimally safe internal temperatures do not hurt protein quality, but grinding meat or cooking with direct high heat can, especially with overcooking. For eggs the direction reversed, with scrambled at 137 against boiled at 135. Nobody should reorganise their cooking around 20 DIAAS points, but stop treating a well-done stovetop burger as identical to a medium roast.

Do I need to combine plant proteins at the same meal?

Not within a single meal. That was the original 1970s complementary-protein claim and it has been abandoned, because the free amino acid pool draws on protein eaten across the day. Combining still matters across that day, because cereals run out of lysine first while legumes run out of sulfur amino acids, so the mixture scores higher than either component. A computational optimisation of pea and rice found a maximum DIAAS of 84 at 41 percent rice, a large improvement over rice alone and still short of 100. Eat a varied set of plant proteins across the day and the arithmetic takes care of itself.

The Bottom Line

Protein quality is a real variable with a real measurement behind it, and most of the numbers circulating in fitness content are not that measurement. The values that came out of actual digestibility trials put dairy concentrates and eggs at the top, most meats comfortably above the FAO threshold, soy just under it, beans and peas well under, and cereal grains out of the conversation entirely.

Then the outcome data flattens most of that ranking. At a 30 gram dose, corn, pea, potato, wheat, and a wheat-corn-pea blend each matched milk protein for muscle protein synthesis. Over 10 to 12 weeks at 1.6 to 2.0 g/kg, plant-based and omnivorous diets produced the same lean mass gains. The residual advantage for animal protein in the pooled literature is small, is confined to muscle mass rather than strength, and lives almost entirely in non-soy plant sources.

What that leaves is a short list of things worth acting on. Hit your total intake first, because it dominates everything else. Anchor on eggs and dairy, which are simultaneously the highest scoring and among the cheapest sources available in 2026. Use legumes for cheap volume and pair them with grains across the day rather than within a meal. Recalculate your cost per 30 grams once a year, because the prices move more than the science does. Skip the grains-and-nuts protein accounting, skip collagen for muscle purposes, and skip anything sold on a DIAAS claim you cannot trace to a study.

The ranking is real. It is also the third most important thing on the list, behind eating enough and eating consistently.

References

1. Mathai, J.K., Liu, Y., & Stein, H.H. (2017). Values for digestible indispensable amino acid scores (DIAAS) for some dairy and plant proteins may better describe protein quality than values calculated using the concept for protein digestibility-corrected amino acid scores (PDCAAS). British Journal of Nutrition, 117(4), 490-499.
2. Bailey, H.M., Mathai, J.K., Berg, E.P., & Stein, H.H. (2020). Most meat products have digestible indispensable amino acid scores that are greater than 100, but processing may increase or reduce protein quality. British Journal of Nutrition, 124(1), 14-22.
3. Fanelli, N.S., Martins, J.C.F.R., & Stein, H.H. (2024). The digestible indispensable amino acid score (DIAAS) in eggs and egg-containing breakfast meals is greater than in toast breads or hash browns served without eggs. Journal of Nutritional Science, 13, e68.
4. Han, F., Moughan, P.J., Li, J., & Pang, S. (2020). Digestible indispensable amino acid scores (DIAAS) of six cooked Chinese pulses. Nutrients, 12(12), 3831.
5. Han, F., Han, F., Wang, Y., Fan, L., Song, G., Chen, X., et al. (2019). Digestible indispensable amino acid scores of nine cooked cereal grains. British Journal of Nutrition, 121(1), 30-41.
6. Herreman, L., Nommensen, P., Pennings, B., & Laus, M.C. (2020). Comprehensive overview of the quality of plant- and animal-sourced proteins based on the digestible indispensable amino acid score. Food Science & Nutrition, 8(10), 5379-5391.
7. Nosworthy, M.G., Medina, G., Franczyk, A.J., Neufeld, J., Appah, P., Utioh, A., et al. (2018). Effect of processing on the in vitro and in vivo protein quality of beans. Nutrients, 10(6), 671.
8. Nosworthy, M.G., Franczyk, A., Neufeld, J., & House, J.D. (2023). Impact of processing on the protein quality of hemp. Food Science & Nutrition, 11(11), 7264-7270.
9. Reynaud, Y., Buffiere, C., Cohade, B., Vauris, M., Liebermann, K., Hafnaoui, N., et al. (2021). True ileal amino acid digestibility and digestible indispensable amino acid scores (DIAASs) of plant-based protein foods. Food Chemistry, 338, 128020.
10. Ertl, P., Steinwidder, A., Schonauer, M., Krimberger, K., & Knaus, W. (2016). Net food production of different livestock: a national analysis for Austria including relative occupation of different land categories. Die Bodenkultur, 67(2), 91-103.
11. House, J.D., Hill, K., Neufeld, J., Franczyk, A., & Nosworthy, M.G. (2019). Determination of the protein quality of almonds as assessed by in vitro and in vivo methodologies. Food Science & Nutrition, 7(9), 2932-2938.
12. Gorissen, S.H.M., Crombag, J.J.R., Senden, J.M.G., Waterval, W.A.H., Bierau, J., Verdijk, L.B., & van Loon, L.J.C. (2018). Protein content and amino acid composition of commercially available plant-based protein isolates. Amino Acids, 50(12), 1685-1695.
13. Pinckaers, P.J.M., Kouw, I.W.K., Hendriks, F.K., van Kranenburg, J.M.X., de Groot, L.C.P.G.M., Verdijk, L.B., et al. (2021). No differences in muscle protein synthesis rates following ingestion of wheat protein, milk protein, and their protein blend in healthy, young males. British Journal of Nutrition, 126(12), 1832-1842.
14. Pinckaers, P.J.M., Hendriks, F.K., Hermans, W.J.H., Goessens, J.P.B., Senden, J.M., van Kranenburg, J.M.X., et al. (2022). Potato protein ingestion increases muscle protein synthesis rates at rest and during recovery from exercise in humans. Medicine & Science in Sports & Exercise, 54(9), 1572-1581.
15. Pinckaers, P.J.M., Kouw, I.W.K., Gorissen, S.H.M., Houben, L.H.P., Senden, J.M., Wodzig, W.K.H.W., et al. (2023). The muscle protein synthetic response to the ingestion of a plant-derived protein blend does not differ from an equivalent amount of milk protein in healthy, young males. Journal of Nutrition, 152(12), 2734-2743.
16. Pinckaers, P.J.M., Smeets, J.S.J., Kouw, I.W.K., Goessens, J.P.B., Gijsen, A.P.B., de Groot, L.C.P.G.M., et al. (2024). Muscle protein synthesis rates are similar following the ingestion of pea and milk protein in healthy young males. European Journal of Nutrition, 63(3), 893-904.
17. Pinckaers, P.J.M., Weijzen, M.E.G., Houben, L.H.P., Zorenc, A.H., Kouw, I.W.K., de Groot, L.C.P.G.M., et al. (2024). The muscle protein synthetic response following corn protein ingestion does not differ from milk protein in healthy, young adults. Amino Acids, 56(1), 8.
18. Hevia-Larrain, V., Gualano, B., Longobardi, I., Gil, S., Fernandes, A.L., Costa, L.A.R., et al. (2021). High-protein plant-based diet versus a protein-matched omnivorous diet to support resistance training adaptations. Sports Medicine, 51(6), 1317-1330.
19. Monteyne, A.J., Coelho, M.O.C., Murton, A.J., Abdelrahman, D.R., Dunlop, M.V., Jameson, T.S.O., et al. (2023). Vegan and omnivorous high protein diets support comparable daily myofibrillar protein synthesis rates and skeletal muscle hypertrophy in young adults. Journal of Nutrition, 153(6), 1680-1695.
20. Monteyne, A.J., Coelho, M.O.C., Porter, C., Abdelrahman, D.R., Jameson, T.S.O., Jackman, S.R., et al. (2020). Mycoprotein ingestion stimulates protein synthesis rates to a greater extent than milk protein in rested and exercised skeletal muscle of healthy young men. American Journal of Clinical Nutrition, 112(2), 318-333.
21. Messina, M., Lynch, H., Dickinson, J.M., & Reed, K.E. (2018). No difference between the effects of supplementing with soy protein versus animal protein on gains in muscle mass and strength in response to resistance exercise. International Journal of Sport Nutrition and Exercise Metabolism, 28(6), 674-685.
22. Reid-McCann, R.J., Brennan, S.F., Ward, N.A., Logan, D., McKinley, M.C., & McEvoy, C.T. (2025). Effect of plant versus animal protein on muscle mass, strength, physical performance, and sarcopenia: a systematic review and meta-analysis of randomized controlled trials. Nutrition Reviews, 83(7), e1581-e1603.
23. Lim, M.T., Pan, B.J., Toh, D.W.K., Sutanto, C.N., & Kim, J.E. (2021). Animal protein versus plant protein in supporting lean mass and muscle strength: a systematic review and meta-analysis of randomized controlled trials. Nutrients, 13(2), 661.
24. Volek, J.S., Volk, B.M., Gomez, A.L., Kunces, L.J., Kupchak, B.R., Freidenreich, D.J., et al. (2013). Whey protein supplementation during resistance training augments lean body mass. Journal of the American College of Nutrition, 32(2), 122-135.
25. Lynch, H.M., Buman, M.P., Dickinson, J.M., Ransdell, L.B., Johnston, C.S., & Wharton, C.M. (2020). No significant differences in muscle growth and strength development when consuming soy and whey protein supplements matched for leucine following a 12 week resistance training program in men and women. International Journal of Environmental Research and Public Health, 17(11), 3871.
26. Ciuris, C., Lynch, H.M., Wharton, C., & Johnston, C.S. (2019). A comparison of dietary protein digestibility, based on DIAAS scoring, in vegetarian and non-vegetarian athletes. Nutrients, 11(12), 3016.
27. Morton, R.W., Murphy, K.T., McKellar, S.R., Schoenfeld, B.J., Henselmans, M., Helms, E., et al. (2018). A systematic review, meta-analysis and meta-regression of the effect of protein supplementation on resistance training-induced gains in muscle mass and strength in healthy adults. British Journal of Sports Medicine, 52(6), 376-384.
28. van Vliet, S., Burd, N.A., & van Loon, L.J.C. (2015). The skeletal muscle anabolic response to plant- versus animal-based protein consumption. Journal of Nutrition, 145(9), 1981-1991.
29. Berrazaga, I., Micard, V., Gueugneau, M., & Walrand, S. (2019). The role of the anabolic properties of plant- versus animal-based protein sources in supporting muscle mass maintenance: a critical review. Nutrients, 11(8), 1825.
30. Hammer, L., Moretti, D., Abbuhl-Eng, L., Kandiah, P., Hilaj, N., Portmann, R., & Egger, L. (2023). Mealworm larvae (Tenebrio molitor) and crickets (Acheta domesticus) show high total protein in vitro digestibility and can provide good-to-excellent protein quality. Frontiers in Nutrition, 10, 1150581.
31. FAO (2013). Dietary Protein Quality Evaluation in Human Nutrition. FAO Food and Nutrition Paper 92. Rome: Food and Agriculture Organization of the United Nations.
32. US Bureau of Labor Statistics. Average price data, US city average, June 2026. Series APU0000708111 (eggs), APU0000703112 (ground beef), APU0000FF1101 (boneless chicken breast), APU0000709112 (fresh whole milk).
33. US Department of Agriculture, Agricultural Research Service. FoodData Central. Protein composition values for foods listed in the cost analysis.