Diet · Protein
Almost every protein number you have been given answers a different question
The figure on the food label is a deficiency floor. The figure quoted by lifters is a plateau measured in people who were not dieting. The figure that matters to most people is the one measured in an energy deficit, and it is the one least often quoted. Here are nine measured intakes, each with the population it came from.
The Archive’s position
This is opinion, not a finding. Everything else on this page is the evidence.
Eat one gram of protein per pound of your goal bodyweight, every day, and treat it as the floor rather than the target. If you want to weigh 200 pounds, you eat 200 grams, whether you currently weigh 175 or 260.
Why we hold it
- The RDA is a deficiency floor, not an optimum. It was derived from nitrogen balance in sedentary adults and answers the question of when you stop losing body nitrogen. Almost nobody reading this is trying to merely not waste away.
- The authors of the paper most often used to argue against this recommended the number themselves. The 2018 meta-analysis that produced the 1.62 g/kg plateau says in its own discussion that given the confidence interval reached 2.20, it may be prudent to recommend about 2.2 g/kg/day for people seeking to maximise training-induced gains in fat-free mass. That is one gram per pound, from the source usually quoted against it.
- A second, independent method lands on the same upper figure. Indicator amino acid oxidation, which measures a labelled tracer in breath rather than counting nitrogen, gave an upper 95% bound of 2.2 g/kg/day in trained lifters. Two methods with nothing in common converging on the same ceiling is worth more than either alone.
- Nearly everyone who cares about this is in an energy deficit, and the deficit changes the answer. In the two controlled trials that tested it directly, roughly 2.3 to 2.4 g/kg beat 1.0 to 1.2 g/kg on lean mass by a wide margin, and in one of them the higher protein group gained muscle while losing fat.
- Goal weight is a usable proxy for lean mass. The sports nutrition literature scales protein to fat-free mass, which nobody knows without a scan. Goal weight approximates it, and it stops a person at 300 pounds calculating a target their lean tissue does not justify.
- The cost of being wrong is asymmetric. In healthy people with normal kidneys, four months at 2.6 to 3.3 g/kg changed no marker of renal, hepatic or lipid health. Under-eating protein in a deficit reliably costs lean tissue you then have to rebuild.
- It is a rule a person can actually follow. A number you can compute in your head and hit without a spreadsheet gets followed, and a more precise number that requires a body composition scan does not.
Where this goes beyond the evidence
The central estimate is 1.6 g/kg, not 2.2, and no synthesis has ever identified a plateau at 2.2. The 2.2 figure is the top of a confidence interval on a model that did not reach statistical significance, which is a reason to doubt the ceiling rather than proof of a higher one. The sharper problem is a trial that tested this directly: 39 adults at a 40% deficit were given 0.8, 1.6 or 2.4 g/kg for three weeks, and while both higher arms beat the RDA on preserving fat-free mass, 1.6 and 2.4 could not be told apart. That is the closest thing to a head-to-head test of this rule, and it found no advantage to the higher number. The honest summary is that the gap between 1.6 and one gram per pound is roughly 40 to 60 grams a day at an intake where no harm has been measured in healthy kidneys. Choosing the top of the interval buys insurance against real uncertainty. It does not buy a measured additional effect, because in the one trial designed to detect one, there was not one to detect.
Who this is not for
Anyone with reduced kidney function or chronic kidney disease. The meta-analysis clearing high protein on kidney function enrolled only adults without kidney disease and its authors limit the conclusion to them, so it does not transfer. The rule also breaks at the edges of its own arithmetic: someone at 150 pounds aiming for 200 would be eating close to 3 g/kg of what they currently weigh, past every measured plateau, and someone very heavy aiming far down lands below the range the deficit literature recommends. It ignores height entirely. And 200 grams is 800 calories, which crowds a smaller person's deficit hard. It is a good default, not a law of physics.
The arithmetic, worked
Take someone who weighs 230 pounds and wants to weigh 200. The rule says 200 grams of protein a day, not 230. In the units the research uses, 200 grams against a goal weight of 90.7 kg is 2.2 g/kg of goal bodyweight.
Now put that against the recommendation for exactly this situation, which is an adult in a deficit trying to hold lean mass. That literature recommends 2.3 to 3.1 grams per kilogram of fat-free mass. If the person in the example carries 160 pounds of lean tissue, 200 grams works out at 2.75 g/kg of fat-free mass, which sits inside that range rather than above it.
That is the case for the rule in one paragraph. It is not that one gram per pound is a magic number. It is that scaling to goal weight lands a person close to the published recommendation for their situation, using a number they can work out without a body composition scan.
What gets dropped when the 1.6 figure is quoted
The 2018 meta-analysis in the British Journal of Sports Medicine is the most cited paper in this argument, and it is a good paper: 49 randomised trials, 1,863 participants. It reported that protein intake above 1.62 g/kg/day produced no further gain in fat-free mass during resistance training. That sentence is usually where the quoting stops.
Two things sit in the paper that rarely travel with it. The breakpoint carried a p value of 0.079, and the authors state in the results that they present it as a segmental regression despite not being statistically significant. And its 95% confidence interval runs from 1.03 to 2.20 g/kg/day. The same dataset is consistent with the plateau sitting near 1 gram per kilogram and consistent with it sitting at 2.2, which is one gram per pound.
This does not make the figure wrong or the paper weak. It makes 1.62 a central estimate with real uncertainty around it rather than a hard ceiling, and anyone using it to say that higher intakes are definitively pointless is claiming more precision than the analysis produced. A third omission matters just as much: every trial in it was conducted in people training at or near maintenance, so it was never a test of what happens when you are dieting.
And then there is the sentence in the paper that almost never gets quoted at all. In the discussion, immediately after reporting the interval, the authors write that given the confidence interval spanned 1.03 to 2.20, it may be prudent to recommend around 2.2 g/kg/day for those seeking to maximise resistance training-induced gains in fat-free mass. The paper universally cited to argue that 1.6 is the ceiling recommends 2.2 in its own text, and 2.2 g/kg is one gram per pound.
A second line of evidence reaches the same number by a completely different route. Indicator amino acid oxidation measures a labelled tracer in exhaled breath instead of counting nitrogen in and out, and in trained lifters it returned an average requirement of 1.7 g/kg/day with an upper bound of 2.2. Two methods sharing no assumptions landing on the same ceiling is a stronger signal than either on its own.
The case against going that high is a single trial, and it is the best one here. Thirty-nine adults at a 40% deficit were assigned 0.8, 1.6 or 2.4 g/kg for three weeks. Both higher arms protected fat-free mass better than the RDA. Between 1.6 and 2.4 there was nothing to separate them. If you want the strongest argument that one gram per pound is more than you need, that is it, and it is on this page rather than left out of it.
Nine measured intakes, and who they were measured in
- 0.83 g/kg/day0.38 g/lb
Measured in
235 individual adults across 19 nitrogen balance studies, pooled to set the current RDA
The 97.5th percentile of the requirement distribution came out at 0.83 g/kg/day, with the median requirement at 0.66 g/kg/day. That is the number printed on food labels.
ReviewMeta-analysis of nitrogen balance studies for estimating protein requirements in healthy adults (2003)What this does not show
This is the intake at which a sedentary adult stops losing body nitrogen. It was never an estimate of the intake that builds or best preserves muscle, and the authors did not present it as one. A floor is not a target.
- 1.62 g/kg/day0.73 g/lb
Measured in
1,863 participants across 49 randomised trials of resistance training, with the breakpoint itself fitted on 42 study arms and 723 participants
The widely quoted plateau. Above roughly 1.6 g/kg/day, added protein produced no further gain in fat-free mass during resistance training.
ReviewA 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 (2018)What this does not show
Two things, and both are routinely dropped when this number is quoted. The breakpoint was not statistically significant, at p = 0.079, and the authors say so in the paper. Its 95% confidence interval runs from 1.03 to 2.20 g/kg/day, so the same data are consistent with a plateau near 1 g/kg and equally consistent with one at 2.2. It was also measured in people training at maintenance, not in people in a deficit.
- 2.4 g/kg/day against 1.2 g/kg/day1.09 g/lb against 0.55 g/lb
Measured in
40 young men, 20 per arm, in a 4 week trial at a 40% energy deficit with six training days a week, body composition by four-compartment model
The high protein group gained 1.2 kg of lean mass while in a deficit. The lower protein group gained 0.1 kg. The high protein group also lost more fat, 4.8 kg against 3.5 kg.
HumanHigher compared with lower dietary protein during an energy deficit combined with intense exercise promotes greater lean mass gain and fat mass loss: a randomized trial (2016)What this does not show
Four weeks, young men, supervised food, and a training load most people will not match. It shows that the deficit changes the answer, not that 2.4 g/kg is the optimum or that the result holds for a year.
- 2.3 g/kg/day against 1.0 g/kg/day1.05 g/lb against 0.45 g/lb
Measured in
20 resistance-trained athletes, 10 per arm, two weeks at 60% of habitual energy intake
Lean mass loss was 0.3 kg on the high protein diet and 1.6 kg on the control. Total weight loss was 1.5 kg against 3.0 kg, meaning most of the extra weight the control group lost was lean tissue.
HumanIncreased protein intake reduces lean body mass loss during weight loss in athletes (2010)What this does not show
Two weeks and 20 people. It also found higher fatigue ratings on the high protein arm, which is reported here because it was in the results.
- 0.8 against 1.6 against 2.4 g/kg/day0.36 against 0.73 against 1.09 g/lb
Measured in
39 adults, randomised to three arms, 10 days at maintenance followed by 21 days at a 40% energy deficit
The only trial that put a middle arm between the two numbers people argue about. Both 1.6 and 2.4 g/kg protected fat-free mass better than the RDA, and preserved the muscle response to a protein-rich meal that was blunted at 0.8. Everyone lost the same 3.2 kg of bodyweight regardless of protein; what differed was what that weight was made of.
HumanEffects of high-protein diets on fat-free mass and muscle protein synthesis following weight loss: a randomized controlled trial (2013)What this does not show
It also found no detectable difference between 1.6 and 2.4. This is the most direct test that exists of whether going above roughly 1.6 g/kg in a deficit buys anything further, and it did not find that it does. Twenty-one days, 39 people.
- 2.2 g/kg/day as the upper 95% confidence bound1.00 g/lb
Measured in
8 male bodybuilders with three or more years of training, 42 separate tracer experiments across intakes from 0.1 to 3.5 g/kg/day, on a rest day
A different method entirely. Indicator amino acid oxidation tracks the oxidation of a labelled amino acid in breath rather than counting nitrogen in and out, and it returned an average requirement of 1.7 g/kg/day with an upper 95% bound of 2.2. That is roughly 2.6 times the current recommendation, measured on a day these men did not train.
HumanIndicator amino acid-derived estimate of dietary protein requirement for male bodybuilders on a nontraining day is several-fold greater than the current recommended dietary allowance (2017)What this does not show
Eight men, one rest day, and a requirement estimate rather than a measured effect on muscle. Its value is that it is methodologically independent of everything else on this page and lands on the same upper figure.
- 2.3 to 3.1 g/kg of fat-free mass per day1.05 to 1.41 g/lb of fat-free mass
Measured in
Six trials in energy-restricted resistance-trained adults with lower body fat, 13 study groups reviewed
The recommendation the authors landed on, scaled upward with how severe the deficit is and how lean the person already is. The one group that was lean, genuinely restricted, and still held its fat-free mass was the group eating the most protein in the review.
ReviewA systematic review of dietary protein during caloric restriction in resistance trained lean athletes: a case for higher intakes (2014)What this does not show
This is expressed per kilogram of fat-free mass, not per kilogram of bodyweight, and the distinction is the whole point. It was derived in lean trained athletes, which is not most people reading it.
- 2.6 to 3.3 g/kg/day for four months1.18 to 1.50 g/lb
Measured in
12 resistance-trained men, randomised crossover, two 8 week periods, full metabolic panel and lipids
No change in blood lipids and no change in markers of kidney or liver function at more than four times the RDA. No change in body composition or performance either.
HumanThe effects of a high protein diet on indices of health and body composition: a crossover trial in resistance-trained men (2016)What this does not show
Twelve healthy young men with normal kidneys, for four months. It is evidence against acute harm in that population. It is not evidence about people who already have kidney disease, and it is not a long-term study.
- 1.5 g/kg/day or more, against lower intakes0.68 g/lb or more
Measured in
1,358 participants across 28 randomised trials in adults without kidney disease, pooled
The kidney question, answered at the scale it deserves. Change in glomerular filtration rate did not differ between higher and lower protein intakes, with a standardised mean difference of 0.11, 95% CI minus 0.05 to 0.27, p = 0.16. Post-intervention GFR was slightly higher on high protein, which the authors read as a normal functional adjustment rather than injury.
ReviewChanges in kidney function do not differ between healthy adults consuming higher- compared with lower- or normal-protein diets: a systematic review and meta-analysis (2018)What this does not show
Every trial in it was run in adults without kidney disease, which is the population the conclusion is limited to. It says nothing about anyone with existing renal impairment, and the authors note an unclear risk of selection bias in the included trials.
Questions people ask about this
- Is one gram per pound of bodyweight the same as one gram per pound of goal weight?
- No, and the difference is the entire point. The common version of the rule scales to what you weigh now, which means a person carrying a lot of fat calculates a target their lean tissue does not justify. Scaling to goal weight approximates lean mass, which is what the sports nutrition literature actually scales to.
- Does more protein than 1.6 g/kg do anything?
- For building muscle at maintenance, the meta-analysis found no further gain above roughly that figure, though the breakpoint was not statistically significant, its confidence interval ran to 2.2 g/kg, and its authors recommended about 2.2 in the discussion. In a deficit, 2.3 to 2.4 g/kg clearly beat 1.0 to 1.2 g/kg. The one trial with a middle arm found 1.6 and 2.4 indistinguishable, so the gap between them is genuinely unsettled rather than decided.
- Will a high protein diet damage my kidneys?
- In people with healthy kidneys there is no trial evidence that it does. A meta-analysis of 28 randomised trials in 1,358 adults without kidney disease found no difference in the change in glomerular filtration rate between higher and lower protein intakes. That evidence is explicitly limited to people without kidney disease and does not transfer to anyone who already has reduced kidney function, where protein load is a clinical decision.
- Why is the RDA so much lower than any of these numbers?
- Because it answers a different question. The RDA was derived from nitrogen balance studies and represents the intake at which 97.5% of sedentary adults stop losing body nitrogen. It is the floor for avoiding deficiency, not an estimate of the intake that best builds or preserves muscle, and the authors never presented it as one.
- Does it matter how I spread protein across the day?
- It matters less than the daily total, which is the thing to get right first. The distribution research is covered separately once the trials have been checked.
Where this sits in the protocol
Protein is the floor the rest of it stands on. It is the variable that decides whether weight lost is fat or muscle, which is the difference between the two outcomes people call the same thing.