Training to Failure vs. Leaving Reps in Reserve: What the Evidence Says About Hypertrophy
One of the longest-running debates in strength training is how close to muscular failure a set needs to go to build muscle. A 2024 series of meta-regressions published in Sports Medicine — led by Zac P. Robinson and colleagues, including well-known resistance-training researchers Jacob Remmert, Martin Refalo, and Michael Zourdos — pooled data across resistance-training studies to model the dose-response relationship between proximity to failure (measured in repetitions in reserve, or RIR) and two separate outcomes: strength gain and muscle hypertrophy. The headline finding is that these two adaptations do not respond to proximity-to-failure the same way.
The mechanistic basis
Muscle growth is driven largely by mechanical tension on muscle fibers, accumulated metabolic stress, and sufficient recruitment of higher-threshold motor units. As a set progresses and fatigue builds, the nervous system recruits progressively more motor units to maintain force output — recruitment that is largely complete only in the last few repetitions before failure. This is the mechanistic argument for training close to or at failure: it maximizes fiber recruitment and time under tension in the "hardest" reps. Strength, by contrast, depends heavily on neural adaptations (rate coding, motor learning, coordination) that can be trained productively without pushing every set to its absolute limit, and excessive failure training generates fatigue that can blunt the quality of subsequent sets and sessions.
What the research shows
- Hypertrophy improves closer to failure. The meta-regression found a negative marginal slope for RIR on muscle size — meaning that, on average, sets terminated nearer to failure produced greater hypertrophic outcomes across the pooled studies.
- Strength gains are largely insensitive to proximity to failure. For strength, the confidence intervals around the marginal slope for RIR included a null value, indicating no meaningful relationship — training with several reps left in the tank produced strength gains statistically comparable to training to or near failure.
- The authors' own summary: "Strength gains were similar across a wide range of RIR, while muscle hypertrophy improves as sets are terminated closer to failure."
- This finding is broadly consistent with a 2024 randomized trial in the Journal of Sports Sciences (Refalo et al.), which compared training to momentary muscular failure against training with 1–3 RIR in resistance-trained adults over eight weeks using a unilateral leg-training model, and reported similar hypertrophy between the two approaches.
- Across the broader literature, outcomes in the roughly 0–4 RIR range tend to cluster closely together, while training much further from failure (very low effort, high RIR) shows clearer decrements in hypertrophic response.
Practical recommendations
- For hypertrophy-focused training: keep most working sets in the 0–3 RIR range (i.e., stopping with zero to three reps left before technical failure). Occasional true failure sets are not harmful and may be useful for select isolation exercises, but are not required on every set.
- For strength-focused training: 2–4 RIR is generally sufficient on compound lifts; training this way preserves bar-speed quality, reduces systemic fatigue, and allows higher training frequency and volume without needing to grind failure reps that add fatigue without adding proportional benefit.
- Manage fatigue deliberately. Failure training — especially on multi-joint compound lifts like squats, deadlifts, and heavy presses — disproportionately increases neuromuscular fatigue and recovery time relative to the extra stimulus it provides. Reserving true failure for machine or single-joint exercises (leg extensions, curls, lateral raises) limits the fatigue cost while still allowing a higher effort ceiling where it's safer to apply.
- Who should be cautious with failure training: beginners still learning technique, anyone training free-weight compound lifts without a spotter, and athletes in-season who need to preserve capacity for sport-specific work should lean toward the RIR-based approach (1–3 RIR) rather than habitual failure training.
- Track proximity to failure consistently. RIR/RPE scales are subjective but become more reliable with practice — logging RIR estimates alongside load and reps over time helps ensure sets are neither too easy nor unnecessarily maximal.
Overall, the current evidence suggests the "always train to failure" and "always stay several reps shy of failure" camps are both oversimplifications. For muscle growth, edging closer to failure appears to matter; for strength, effort proximity matters far less than consistent, progressive loading — meaning the smartest approach differs depending on the primary training goal.
Helpful products
- Adjustable dumbbell set — useful for fine-tuning load so you can dial in precise RIR/RPE targets on isolation and accessory lifts.
- Strength training log book — helps track RIR/RPE estimates, load, and reps across sessions to monitor progressive overload.
