How much protein do you need to build muscle?

In short

The official RDA of 0.8 g/kg/day is the amount needed to avoid deficiency — it was never intended as the amount that maximises muscle growth.
Two large dose–response analyses place the point of diminishing returns at 1.3 g/kg/day and 1.62 g/kg/day. The 95% confidence interval on the latter runs from 1.03 to 2.20.
Extra protein does almost nothing on its own. In the largest analysis of healthy adults, the effect on lean mass was statistically significant only when resistance training was part of the intervention.
The measured effect size is real but modest: roughly 0.3 kg of additional fat-free mass over a typical trial.

The number most people have heard is answering a different question

The Recommended Dietary Allowance for protein in the United States and Canada is 0.8 g per kg of body mass per day.⁵ It is worth being precise about what that figure means: an RDA is set to meet the requirements of 97–98% of healthy people, and the requirement in question is nitrogen balance — not losing body protein. It is a floor for adequacy, not a target for adaptation.

Asking how much protein maximises the muscle you gain from lifting is a different question, and it has been studied directly. Because individual trials in this area are small and noisy, the useful evidence comes from meta-regression: pooling many randomised trials and modelling gains against total daily protein intake to find where the curve flattens.

What the dose–response analyses found

Three pooled analyses dominate the current literature. They differ in scope, and reading them together is more informative than reading any one alone.

Table 1. Pooled analyses of protein intake and lean mass in healthy adults. FFM = fat-free mass; LBM = lean body mass; SMD = standardised mean difference; RT = resistance training.
Analysis	Trials / participants	Headline result
Morton et al., 2018 Br J Sports Med	49 RCTs 1,863	Supplementation beyond a total intake of 1.62 g/kg/day produced no further gains in FFM (95% CI 1.03–2.20). Overall effect: +0.30 kg FFM (0.09–0.52) and +2.49 kg on 1RM strength (0.64–4.33).
Tagawa et al., 2021 Nutrition Reviews	105 RCTs 5,402	Each additional 0.1 g/kg/day was associated with +0.39 kg LBM below 1.3 g/kg/day (95% CI 0.36–0.41) but only +0.12 kg above it (0.11–0.14). Above that threshold the effect kept rising with RT and declined without it.
Nunes et al., 2022 J Cachexia Sarcopenia Muscle	74 RCTs 2,665	Small but significant effect on LBM overall (SMD 0.22, 95% CI 0.15–0.30) and with RT (SMD 0.22, 0.14–0.30). Without resistance exercise the effect was not significant (SMD 0.21, −0.15 to 0.58; 6 studies).

Figure 1. Where the gains stop. The blue bar is the 95% confidence interval Morton and colleagues reported around their 1.62 g/kg/day breakpoint; it spans 1.03 to 2.20, which is most of the plausible range people argue about. Tagawa and colleagues did not report a confidence interval on their 1.3 g/kg/day inflection point.

Why the two headline numbers disagree

1.3 and 1.62 are often quoted as if one must be wrong. They are better read as two estimates of the same soft edge, produced by analyses that asked slightly different questions.

Morton and colleagues restricted their analysis to supplementation trials layered on top of resistance training and looked for the intake beyond which supplementing added nothing. Tagawa and colleagues pooled a much larger and more heterogeneous set — 105 trials, only about half of which involved resistance training — and modelled the shape of the whole curve rather than a single cut point. Their finding that the curve keeps climbing above 1.3 g/kg/day in the training subgroup but falls away without training is the reconciliation: the plateau is later when you are actually lifting.

The honest summary is not a single number. It is that the return on each additional gram falls off steeply somewhere in the region of 1.3–1.6 g/kg/day, and that the data cannot resolve it more finely than that. Morton's confidence interval — 1.03 to 2.20 — is a candid statement of how much room remains.

Protein does very little without the training stimulus

This is the most consistently under-reported finding in the set. Nunes and colleagues separated trials that paired extra protein with resistance exercise from the handful that did not. With resistance exercise, the effect on lean body mass was small but statistically robust. Without it, across six studies, the confidence interval crossed zero (SMD 0.21, 95% CI −0.15 to 0.58).

Tagawa's group found the same asymmetry in the shape of the curve: past 1.3 g/kg/day, the effect on lean mass continued to rise in the trials with resistance training and declined in those without. Protein is permissive, not causal. It supplies the substrate; the training supplies the signal.

It is also worth keeping the effect size in view. Morton's pooled estimate was +0.30 kg of fat-free mass, with a lower confidence bound of 0.09 kg. Nunes and colleagues translated their standardised effect to roughly a 0.5–0.7 kg difference between groups. These are real, replicated effects, and they are small ones. Protein intake is a lever worth pulling once, not a lever worth obsessing over.

Does it matter how you spread it across the day?

Whole-day totals dominate, but per-meal dosing has a defensible rationale: the synthetic response to a single feeding saturates, so a day's protein taken in one sitting is unlikely to be used as efficiently as the same amount spread out.

Schoenfeld and Aragon reviewed this literature and proposed a distribution rule rather than a hard per-meal cap. Their recommendation, verbatim: consume protein at a target intake of 0.4 g/kg/meal across a minimum of four meals in order to reach a minimum of 1.6 g/kg/day. They put the practical upper bound of a usefully large meal at 0.55 g/kg.

Note what this construction does. Four meals at 0.4 g/kg lands you on 1.6 g/kg/day — the same neighbourhood the dose–response analyses point to, reached from the opposite direction. Two independent lines of reasoning arriving at the same place is the strongest thing this literature has going for it.

Table 2. The 0.4 g/kg/meal × 4 meals rule, converted to grams. Daily totals are the sum of four such meals, equivalent to 1.6 g/kg/day.
Body mass	Per meal (0.4 g/kg)	Per day (1.6 g/kg)
60 kg / 132 lb	24 g	96 g
70 kg / 154 lb	28 g	112 g
80 kg / 176 lb	32 g	128 g
90 kg / 198 lb	36 g	144 g
100 kg / 220 lb	40 g	160 g

Two qualifications worth carrying

Age shifts the target upward. Morton's meta-regression found the benefit of supplementation shrank as participant age rose (−0.01 kg per year, 95% CI −0.02 to −0.00, p = 0.002), which the authors read as evidence that older adults need higher intakes to get the same response. Nunes and colleagues found the significant effect in adults under 65 clustered at intakes of 1.6 g/kg/day and above.

Most trial participants were not starting from zero. Nunes and colleagues noted that around 80% of the studies they pooled reported baseline intakes of at least 1.2 g/kg BW — already 50% above the RDA. The measured effects are therefore effects of going from fairly high to higher, not from deficient to sufficient. If your current intake is genuinely low, the trials above almost certainly understate what you would gain by fixing it.

The bottom line

If you are lifting and want to build muscle, aim for roughly 1.6 g of protein per kg of body mass per day, spread across four or more meals of about 0.4 g/kg each. Below that range the evidence says you are leaving gains on the table; above it, the return on each extra gram is small enough that adherence, total energy intake and the training itself matter considerably more.

Getting there is mostly a measurement problem rather than a knowledge problem — few people can estimate a day's protein without writing it down. Kettle logs food macros, including protein, alongside your training, so the number and the stimulus it is meant to support live in the same place.