Stopping resistance-training sets a few reps short of failure produces similar strength gains to training to failure, with faster recovery
Interpretation
Strength: consistent across meta-analyses (Grgic 2022, 15 studies in young adults, ES -0.09; Robinson 2024 meta-regressions, strength similar across a wide range of reps in reserve). Hypertrophy shows a small advantage closer to failure (Refalo 2023 ES 0.19; Robinson 2024). Fatigue: one small crossover in 10 resistance-trained men showed slower 24-48 h recovery after failure than after equal volume short of failure (Morán-Navarro 2017), and 22 young men stopping at 20% velocity loss matched strength gains and improved jump more than at 40% (Pareja-Blanco 2017). People underestimate reps in reserve by about one (Halperin 2022, 414 participants). Graded moderate: the strength finding is well replicated, but the recovery evidence is small, all-male, and no study includes endurance athletes or measures next-day endurance performance directly (Doma 2017 review supports carry-over fatigue generally).
Practical implication
Triathletes can stop most sets with about 1 to 3 reps in reserve, keep heavy compound lifts furthest from failure and adjust load on tired days, expecting similar strength gains with less carry-over fatigue. The recovery benefit for swim, bike and run sessions is extrapolated from lifters, not tested in triathletes.
Sources
8 sources, independently verified.
No significant difference between failure and non-failure training for strength (ES -0.09, 95% CI -0.22 to 0.05) or hypertrophy (ES 0.22, -0.11 to 0.55); small hypertrophy advantage for failure in resistance-trained people (ES 0.15). Young adults only. Published online 2021.
Exploratory meta-regressions with reps in reserve estimated from study descriptions: strength gains were similar across a wide range of RIR (slopes' CIs included zero); hypertrophy increased as sets ended closer to failure (negative slopes, CIs excluding zero). Modest model fit.
Training to set failure had a trivial hypertrophy advantage (ES 0.19, 95% CI 0.00-0.37); momentary failure vs non-failure not significant (ES 0.12, -0.13 to 0.37); higher velocity-loss thresholds did not consistently increase growth (ES 0.08). Suggests a possibly non-linear relationship. Published online 2022.
Bench press and full squat: 3 x 5 and 6 x 5 with a 10-repetition load (non-failure) vs 3 x 10 to failure. Failure caused larger acute drops in countermovement jump and lifting velocity, even vs the equal-volume 6 x 5, and recovery between 24 and 48 h was faster after the non-failure protocols; CK, ammonia and growth hormone followed markedly different courses. Small, men only.
Eight weeks of squats stopping sets at 20% or 40% velocity loss: similar strength gains; countermovement jump improved more with 20% (9.5% vs 3.5%) despite fewer repetitions; 40% produced more vastus intermedius hypertrophy and reduced type IIX fibres, which 20% preserved. Published online 2016.
People underpredicted repetitions to failure by 0.95 reps on average (95% CI 0.17-1.73; I² 97.9%). Accuracy better with heavier loads (≤12 reps), slightly better closer to failure and in later sets; training status had no clear effect. Published online 2021.
Introduced an RPE scale anchored to reps in reserve (10 = 0 RIR, 9 = 1 RIR). Bar velocity was strongly inversely related to RPE in experienced (r = -0.88) and novice (r = -0.77) squatters. Sex not reported.
Review of studies measuring endurance performance after single resistance sessions: residual fatigue can impair a subsequent endurance session for several hours to several days if recovery is inadequate, via altered neural recruitment, reduced movement efficiency, muscle soreness and lower glycogen. Recommends adjusting session order, recovery between modes and intensity. Narrative synthesis, not a meta-analysis.