Loading fundamentals 01: Rest between sets

Short answer. Rest length appears to matter for growth mainly because it changes how many repetitions you complete on later sets: when one study equated volume load, the difference disappeared. Rest until your next set is not compromised, around two minutes on most exercises. If you are short on time, cut a set or pair non-competing exercises, and leave the rest alone.

What it is

The dominant recommendation for hypertrophy has been short rest. The textbook of the National Strength and Conditioning Association (NSCA) prescribes 30 to 90 seconds (Haff and Triplett, 2015) and most commercial programs inherit it.

That number came from a hypothesis about hormones. The hypothesis has since been tested directly and it failed. The number is still in circulation.

The proposed mechanism

Short rest was recommended because of hormones. Acute research showed that compressed rest periods amplify the post-exercise elevation in testosterone and growth hormone, and that elevation was theorized to drive muscular adaptation (Kraemer and Ratamess, 2005). The prescription follows cleanly from the theory: rest less, spike more, grow more.

The theory has been tested directly since. In resistance-trained young men, neither load nor systemic hormone elevations determined training-mediated hypertrophy or strength gains (Morton et al., 2016). A follow-up from the same group found that muscle androgen receptor content, not circulating hormones, was associated with hypertrophy (Morton et al., 2018).

The rationale for short rest has been removed. The recommendation has not. That gap is what this piece is about.

There is a competing mechanism, and it runs the other way. Shorter rest reduces the number of repetitions you complete on subsequent sets at a given load. Since mechanical tension is a primary driver of hypertrophy, fewer repetitions across a session means less of the thing that causes growth. Consistent with this, short rest of one minute blunted the myofibrillar protein synthetic response compared with five minutes, despite producing higher acute testosterone (McKendry et al., 2016).

What has actually been tested

A Bayesian meta-analysis pooled nine studies with 19 hypertrophy measurements, comparing rest durations while holding all other training variables constant (Singer et al., 2024).

Everything worked. Effect sizes for muscle growth, expressed in standard deviations, were medium to large across every rest category. Short rest, at or under 60 seconds, scored 0.48, with a 95% credible interval (the Bayesian range within which the true effect most probably lies) of 0.19 to 0.81; longer rest scored 0.56 (0.24 to 0.86). Split four ways, short 0.47, intermediate 0.65, long 0.55, very long 0.50. The intervals overlap heavily. Nobody in these studies failed to grow because of their rest interval.

The controlled comparison favors longer rest, slightly. For the thigh, the effect favoring longer rest was 0.17 (credible interval -0.13 to 0.43), with an 88% probability the effect exceeded zero but only a 54% probability it exceeded "small." For the arm, 0.13 (-0.27 to 0.51), with a 74% probability above zero. Whole-body measures marginally favored shorter rest at -0.08 (-0.45 to 0.29), which the authors flag as the least reliable of the three because those studies used fat-free mass estimates rather than direct muscle imaging.

More rest stops helping around 90 seconds. The analysis detected no appreciable further benefit beyond that point, consistent with the observation that volume load differences level off between 120 and 180 seconds.

And the most informative study equated volume load and watched the difference disappear. Training the quadriceps with 180-second rest produced greater cross-sectional area gains than 60-second rest across ten weeks, 13.1% against 6.8%. The average repetitions completed across three sets were 16.1 against 9.8. When additional sets were added to equate volume load between conditions, hypertrophy was similar (Longo et al., 2022). The paper's title states the finding plainly: volume load rather than resting interval influences hypertrophy.

Proximity to failure did not change the picture. Whether sets were taken to failure or stopped short, the interaction between rest duration and hypertrophy was unchanged.

The authors' conclusion was unusually direct for a meta-analysis: their findings are inconsistent with the NSCA recommendation of 30 to 90 seconds, and current guidelines warrant reconsideration.

Our reading

Rest interval is not a variable with its own effect on muscle growth. It is a lever on volume load, and volume load is the thing with the effect. Once you see it that way most of the literature resolves.

Short rest looks worse in the data because it costs you repetitions. Long rest looks better because it preserves them. Equate the repetitions and the difference goes away, which is exactly what happened when someone tested it. The plateau at around 90 seconds is where the repetition cost stops accumulating, and there is nothing left for more rest to recover.

So the practical position is that you should rest until your next set is not compromised, and that stopping there costs you nothing. For most people on most exercises that lands somewhere around two minutes, longer for heavy compound work where the fatigue cost is larger, shorter for isolation work on small muscles.

The corollary matters more than the rule. If you are short on time, do not buy it by cutting rest. Cutting rest is the one economy that directly removes the repetitions driving the adaptation. Cutting a set, or pairing exercises that do not compete for the same muscles, leaves the repetitions on the sets you do perform intact. The 60-second circuit is a worse trade than it looks.

The same reasoning applies to where you end a set rather than how long you wait between them: protect the quality of the sets that come after the one you are currently doing. (→ Loading fundamentals 02: Reps in reserve)

Where the goals diverge

Training for strength. Longer rest has stronger justification here than for hypertrophy, because heavy singles and doubles are more sensitive to incomplete recovery, and there is no volume-equating argument to fall back on when the set is three repetitions.

Training for muscular endurance or work capacity. Short rest is a training target in itself rather than a compromise, and this literature does not apply. (→ Loading fundamentals 04: Load and repetition range)

Training with very high set volumes. (→ Loading fundamentals 07: Weekly volume) All the studies here used moderate volumes. The authors note that how rest interacts with high set counts per muscle is undetermined, and there is a plausible argument that fatigue accumulates differently there.

Training in limited time. The trade is real, and the analysis above says which side of it to give up. (→ Loading fundamentals sub-hub)

Managing session fatigue. Rest is one of three levers on how much capacity survives to your later sets. The others are where you terminate each set and how you distribute the week's volume. (→ Loading fundamentals 02: Reps in reserve; Loading fundamentals 03: Training frequency)

The case against

The trained-lifter data barely exists. Six of the nine studies used untrained participants. The authors had insufficient data to run a subanalysis on trained lifters, which means the population most likely to be reading this is the population least represented in it.

The differences are small and the intervals cross zero. A 54% probability that an effect exceeds "small" (Singer et al., 2024) is not a strong finding. We have built a practical position on top of a modest signal, and the fair description of that signal is that longer rest probably helps a little.

The volume-load explanation is a hypothesis that fits, not a demonstrated mechanism. One study equated volume load and the difference vanished. That is one study, with unilateral leg press, in a specific population.

Study durations were five to ten weeks. The authors note that accumulated differences might be more appreciable over longer time frames, which cuts against our reassurance that short rest costs you little.

What is not known

How rest interacts with torso musculature. No study in this analysis examined the chest, back or shoulders directly.

Whether exercise complexity should change the prescription. A heavy squat and a cable curl have different fatigue profiles across sets, and rest has been prescribed identically for both without anyone testing whether it should be.

Whether any of this holds in trained lifters at high volumes over a year. Every answer in this piece is bounded by five to ten weeks of mostly untrained participants.