TLDR: For muscle growth, training to failure produces only a small, borderline advantage over stopping a rep or two short, and for most lifters the difference is not worth the added fatigue. For strength, the picture reverses slightly. Training to failure does not reliably improve strength gains and may modestly work against them. Stopping around one to three reps in reserve on most working sets is a defensible default for most training goals.
What the research says
The most direct evidence comes from a 2023 systematic review and meta-analysis by Martin Refalo and colleagues, published in Sports Medicine, which pooled 15 studies comparing training to momentary muscular failure against various forms of non-failure training. Across studies using any definition of set failure, the analysis found a trivial advantage for training to failure on hypertrophy, with an effect size of 0.19 and a confidence interval that ran from 0.00 to 0.37, technically statistically significant but small enough that its practical importance is limited.
A follow-up 2024 paper by Zac Robinson and colleagues, also in Sports Medicine, took a different approach, modeling proximity to failure as a continuous variable using estimated repetitions in reserve rather than a simple failure versus non-failure split. That analysis found that muscle hypertrophy did increase modestly as sets moved closer to failure, but the relationship was not a straight line. Gains tapered off at the very closest proximities, around zero to one rep in reserve, suggesting a point of diminishing returns rather than a case for maximal effort on every set. Strength told a different story in the same analysis: training closer to failure was associated with slightly smaller strength gains, which the authors attributed to fatigue interfering with the quality of subsequent training.
A randomized trial published by Refalo and colleagues in 2024 in the Journal of Sports Sciences compared training to failure directly against stopping one to two reps short across eight weeks in resistance-trained adults, and found comparable quadriceps growth between the two approaches.
Why the effect is small rather than zero
The mechanistic case for training close to failure rests on motor unit recruitment. As a set progresses and fatigue accumulates, the body recruits higher-threshold motor units, the ones most associated with growth, to keep producing force. Stopping well short of failure, at four or five reps in reserve or more, means many of a set’s later, higher-effort repetitions never happen, and the meta-analytic data does show a meaningful drop-off in hypertrophy at that end of the spectrum.
The picture changes once a set gets close to failure. The marginal repetitions gained by pushing from two reps in reserve to zero come with a disproportionate rise in fatigue and, per the Robinson analysis, without a matching rise in growth. That combination, small upside near failure and a real fatigue cost, is why leaving a rep or two in the tank tends to be recommended over grinding every set to its absolute end.
The practical takeaway
Training close to failure, but not necessarily to failure itself, appears to capture most of the hypertrophy benefit while avoiding the larger fatigue cost and the potential strength penalty of constant maximal effort. One to three reps in reserve on most working sets is a reasonable working range for hypertrophy-focused training. Reserving true failure for occasional sets, rather than every set of every session, is more consistent with the current evidence than either extreme.
This connects to two things covered elsewhere on OUTWRQ. The 2026 ACSM resistance training guidelines explicitly state that training to failure is not required for meaningful strength or hypertrophy gains, which lines up with what the data above shows. And the broader question of where a set stops mattering for growth connects directly to Junk Volume: How Much Training Is Too Much: a set left too far from failure risks becoming volume without much of a stimulus behind it, in the same way that excess sets past a muscle’s weekly ceiling do.
Reps in reserve is also one of the harder things to judge consistently without some kind of record. Most people’s sense of how close they were to failure drifts across a session as fatigue builds. Logging working weight and reps every session, which OUTWRQ does automatically, makes it easier to notice when a set that felt like two reps in reserve was actually closer to five.
Written by the OUTWRQ team. Published August 2026.