Protein on a plant based diet: what the doctors say
Protein on a plant based diet: what McDougall, Esselstyn, Campbell, Barnard and other doctors say, where they agree and differ, with sources.
Updated . For education only, not medical advice.
Our view
The question plant based eaters hear most is where their protein comes from, and the doctors on this site answer it the same way: a varied diet of whole plant foods, eaten in enough quantity, does not run short123. Every plant food carries all the essential amino acids, which is why Pritikin, Novick, Barnard and Gardner all reject the old rule of pairing beans with grains at each meal4567.
We give most weight to the clinicians who have fed patients this way for decades, and for McDougall, Goldhamer and Caldwell Esselstyn the real risk is the opposite one: too much protein, above all from animal foods8910. The shares of calories they name, from about 5 to 10 percent, are reached by beans, grains, potatoes and vegetables without any planning111213.
None of them sees a place for protein powders or shakes: Goldhamer would not give them to anyone, and Esselstyn’s FAQ tells people to leave protein drinks alone1210. What matters more is eating enough food, the point Chef AJ, Rip Esselstyn and Pritikin all make when they tie protein to total calories141516.
Where they agree
A protein shortage is not a practical risk on whole plant foods eaten to appetite: McDougall calls the fear of it a dangerous falsehood, Rogers tells people to stop worrying, and Rip Esselstyn’s site says the problem barely exists unless calories are short11718.
They agree that plant proteins are complete: Pritikin in 1982, Novick in an essay from 2003, Barnard in 2024 and Gardner in 2026 all dismiss the idea that plants lack amino acids or must be combined4567. Pritikin and Novick both trace the combining advice to Diet for a Small Planet, published in 1971, and Novick notes that its author later withdrew it195.
The official allowance, 0.8 grams for each kilogram a person weighs, already holds a wide margin: a dietitian of the Ornish program called it generous in 2016, and Gardner explained in 2026 that it adds two standard deviations to an average need of about 0.662021. A 2025 VegSource video from Jeff Nelson’s channel adds that most people eat about twice the allowance, and Gardner put the usual American intake at about 1.2 grams per kilogram2223.
When they name protein foods, beans, lentils and peas come first for Gardner, Novick and Chef AJ, and Barnard notes that lentils alone would cover the need2425263.
Where they differ
Beyond the shared view that a shortage is unlikely, they differ on the numbers, on whether total protein or its source is the problem, on athletes, and on a few positions that belong to their dates11272829.
How much people need
McDougall put adult needs at about 5 percent of calories in 2021, and Rip Esselstyn gave a range of 4 to 8 percent in 20181115. Campbell named 8 to 10 percent in a lecture posted in 2011, and Goldhamer about 10 percent in 2025, a share he said hardly moves with age3012. Rogers goes lowest: in 2022 he cited a 1982 Lancet paper in which a diet giving only 2.5 percent of calories as protein brought 66 malnourished patients back to health, and in 2025 he called the 0.8 gram allowance far too high3132. In 1982 Pritikin cited Purdue research where rice providing 6 percent of calories as protein was enough for young men, while his own diet gave about 13 percent33.
Total amount or source
McDougall and Rogers want protein low in general: in 2022 McDougall ranked it second only to fat as a harm of Western eating, for its load on liver, kidneys and bone, and Rogers told viewers that year to keep it as low as they can, animal protein above all834. Campbell’s research centers on casein, the main protein of cow’s milk, which in his rat studies turned cancer growth on at higher doses and off at lower ones, while soy and wheat protein did not3536. Ornish, Barnard and Kahleova put the weight on source: Ornish cited a 2014 study tying diets rich in animal protein to higher early mortality, Barnard in 2024 cited Harvard data on lower death rates when plant protein replaced animal protein, and a 2026 trial Kahleova led found weight loss when the swap kept total protein the same273738. In the same vein, Jenkins said in 2026 that soy protein lowers LDL, and Willett said in 2020 that red meat fares badly next to nuts, legumes and soy3940.
Athletes and powders
Novick in 2012 cited a dietetics position paper allowing about 1.2 to 1.7 grams per kilogram for strength athletes, yet argued that ordinary food covers even a heavy lifter and that shakes tip intake into excess2825. Gardner noted in 2026 that people who take in 3,000 to 5,000 calories a day end up near 1.6 grams per kilogram without trying, and cited a trial, paid for by the beef checkoff, where vegans gained as much muscle as omnivores when both ate 1.1 grams per kilogram4142. Goldhamer said in 2025 that active people get more protein simply because they eat more, and he rules out powders for everyone12.
Newer warnings
Gardner, who sat on the US Dietary Guidelines Advisory Committee, objected in 2026 that the guidelines published after his committee’s report urged protein at every meal, while his committee had found too much meat rather than too little protein43. Kahleova said the same year that public advice keeps pushing more protein without saying which kind44. When a channel warned in 2025 that New Zealand vegans fell short on lysine and leucine, the VegSource reply pointed out that nobody in that study was ill45. Gardner grants small gaps, beans a little low in methionine and grains in lysine, but says they stop mattering at the amounts people actually eat746.
Older views
Walter Kempner’s rice diet at Duke, described in a 1975 paper, began very low in protein and later added vegetables and then lean poultry or meat29. In 1982 Pritikin dismissed the protein combining advice of Diet for a Small Planet as unscientific and more harmful than helpful19. Campbell’s own path started in the Philippines, on a project to feed malnourished children more protein, where, he said in 2016, the children eating the most protein seemed the most prone to liver cancer47.
What each expert says
JMJohn McDougall
McDougall builds his case on the experiments of William Rose, who set his recommended amino acid intakes at twice the highest need he found; in McDougall’s 2021 account every starch and vegetable still clears that bar48. He also cites a six-month study in which two athletes lived almost only on potatoes and stayed healthy49. For him the harm comes from too much protein, not too little8.
In a 2012 video Esselstyn said people eating a mix of grains, beans, vegetables and fruit will not lack protein2. His 2014 paper reassured heart patients that varied plant foods meet their amino acid needs, and his FAQ warns against protein drinks, especially those made from animal protein5010.
Campbell’s view of protein grew out of his work on casein, the milk protein that drove cancer growth in his rats35. In a lecture posted in 2011 he put human needs at about 8 to 10 percent of calories, below the 11 to 25 percent many Americans eat, and in 2017 he said whole plant foods supply that much with ease3013.
In a 2024 fact check Barnard said 2,000 calories of broccoli would give three times the official protein amount, and called food combining an outdated idea36. The same year he cited Harvard research favoring plant over animal protein, and in 2026 he noted that soy milk matches cow’s milk at about 8 grams a serving3751.
Ornish argued in 2014 that the protein debate should turn on where it comes from, citing a study on animal protein and early death27. A column on his program’s site, with comments from 2016, found more plant protein tied to no extra risk, and in 2022 he added that producing meat protein uses around 14 times the resources of plant protein5253.
Goldhamer’s 2022 outline of a day puts protein near 10 percent of calories, fat at 15 to 18 percent and complex carbohydrates in the rest54. In 2025 he said that on whole plant foods the risk is excess protein, because beans, nuts and seeds hold so much of it9.
Rogers holds that no natural diet leaves people short of protein, and that intake, animal protein most of all, should stay as low as possible3134. In 2024 he named one exception: a big wound that has to heal after someone has gone days without eating17.
Novick wrote in 2012 that a steady excess of protein may burden the kidneys, liver and bones and raise the hormone IGF-1, and he prefers beans, lentils and peas as sources25. In an essay from 2003 he argued that varied whole plant foods with enough calories can hardly come up short on any amino acid; the only diet he would question is one of fruit alone55.
In 2025 a VegSource video took on another channel’s warning of a protein gap in vegans, noting that the New Zealand study behind the warning found no one ill45. The same video said the official allowances carry deliberate padding22.
In a 2018 talk Rip Esselstyn put protein needs at about 4 to 8 percent of calories, met by anyone who eats enough15. A 2025 article on his site lists greens, whole grains, beans, lentils, tofu, tempeh, nuts and seeds as sources18.
Chef AJ does not track protein; in a 2017 episode she said the problem she meets in her program is people eating too little food overall14. She pointed out that breast milk, made for the fastest growth of life, has protein below a tenth of its calories, and told worried viewers to add beans and other legumes5626.
In a 2026 webinar Gardner explained that no test reveals what a single person needs in protein, and that the allowance works out to about 45 grams daily for women and 55 for men5721. He said animal foods make the target easier to hit but are not needed to reach it58.
On a 2020 panel Willett said that in trials red meat raises LDL compared with plant proteins like nuts, legumes and soy, and that in cohort studies it goes with more heart disease40.
Pritikin told McDougall in 1982 that anyone eating enough to hold their weight gets enough protein, and that even a dietitian would struggle to design a diet short of it16. He named gelatin and egg white, not plant foods, as proteins missing some amino acids4.
In a 2026 study Kahleova led, postmenopausal women lost weight on soybeans, beans and grains in place of animal protein, with no change in how much protein they ate overall38. The release links part of the effect to lower methionine, an amino acid more abundant in meat, eggs and dairy59.
Jenkins said in 2026 that soy protein lowers LDL in two ways: through its own amino acid mix and by taking the place of foods high in saturated fat and cholesterol39.
Kempner’s 1975 paper, covering 106 people at Duke with severe obesity, describes a starting rice diet low in calories, salt, protein and fat, with vegetables usually added after a month or more and lean poultry or meat later still29.
We eat a varied diet of whole plant foods, with beans, lentils, grains, potatoes and vegetables, and enough of them to hold a healthy weight, as these doctors describe216. We do not count protein, and we leave protein powders and shakes on the shelf. How much protein a particular person needs, during pregnancy, recovery or illness for instance, is a question for your own doctor. This page is education, not medical advice.
Each small number links to its reference below: the video at the exact second, or the page where it was said.