How Close to Failure Should You Train for Muscle?
By Mac Burke, qualified sports nutritionist
The short answer
For muscle growth, finish most sets with about 0-3 reps left in reserve. Research shows going all the way to failure gives at most a small extra benefit, while adding fatigue. Keep true failure for the last set of machine or isolation exercises, and stay 1-3 reps short on heavy squats, deadlifts and presses.
Hard sets build muscle. Easy sets do not do much. But hard does not have to mean failure every time. The evidence points to a zone: stop most sets when you have about 0-3 good reps left. That gets nearly all the growth with less fatigue, less injury risk and better performance on the sets that follow.
Is training to failure better for muscle growth?
Only slightly, if at all. A 2023 meta-analysis of 15 studies found a trivial advantage for training to set failure over stopping short (effect size 0.19). When the analysis looked only at true momentary failure versus non-failure, there was no clear advantage (Refalo et al., 2023).
A 2022 meta-analysis, also of 15 studies, found no significant overall difference for strength or size. It did find a small benefit for failure training in trained lifters for hypertrophy (effect size 0.15) (Grgic et al., 2022). So failure is not required, but it is not harmful to growth either.
Does growth drop off the further you stop from failure?
It seems to. A 2024 series of meta-regressions found muscle growth improved as sets ended closer to failure, while strength gains showed little link to proximity to failure (Robinson et al., 2024). The authors called it exploratory because reps in reserve were estimated, not measured.
Put the studies together and you get a practical rule. Stopping 5-6 reps short likely leaves growth on the table. Stopping 0-3 reps short gets you most of it. Grinding to failure on every set adds little and costs a lot.
What does reps in reserve (RIR) mean in practice?
RIR is how many more clean reps you could have done when you racked the weight. A set of 10 with 2 RIR means you could have done 12.
- 0 RIR: you could not do another rep with good form. This is failure.
- 1 RIR: one more rep was possible, but it would have been slow and ugly.
- 2-3 RIR: the last reps slowed down clearly. This is the main working zone.
- 4+ RIR: the set felt moderate. Fine for warm-ups and technique, light on growth.
Most people underestimate how many reps they have left, especially on higher-rep sets. A good test is to take one set of a safe exercise, like a machine row, to true failure. If you guessed 2 RIR at rep 10 and then hit 15, your guesses are too cautious.
Here is how that looks on a working day. Say you do 3 sets of 10 on the leg press. Set one ends with about 3 reps left. Set two ends with 2 left. The last set goes to 0-1. You still collect hard sets, you finish fresh enough for the next exercise, and only one set carries the full cost of failure.
Does failure work differently for heavy and light loads?
In Refalo's analysis, the small advantage for failure did not depend on whether loads were light or heavy (Refalo et al., 2023). In practice, though, light loads need more care. On a set of 6, stopping 2 reps short is easy to judge. On a set of 25, the burn and heavy breathing make most people stop far earlier than they think.
So the lighter the load, the closer to failure you should push. If you use high-rep sets on isolation work, take the last set to failure. If you use heavy sets of 5-8 on barbell lifts, stopping 1-3 reps short gives you nearly everything with far less risk.
Which exercises are safe to take to failure?
Match the effort to the risk. Failing a leg extension is harmless. Failing a heavy squat without safeties is not.
- Take to 0-1 RIR: machines, cables, curls, lateral raises, calf raises, leg curls.
- Keep at 1-3 RIR: barbell squats, deadlifts, bench press, overhead press, heavy rows.
- Use failure on the last set of an exercise, not every set, so the earlier sets keep their quality.
Can training to failure slow recovery?
Yes. Going to failure leaves you more tired, and that fatigue can lower your reps on the next sets and the next session. The 2022 meta-analysis found that when volume was not matched, stopping short of failure actually produced better strength gains (effect size -0.32), likely because people could do more good work (Grgic et al., 2022).
Strength seems to care even less about failure. The 2024 meta-regressions found little link between how close sets ended to failure and strength gains (Robinson et al., 2024). If you want both size and strength, keep your heavy lifts a little further from failure and push your accessory work harder.
For a busy adult, that matters. Work stress, poor sleep and long days already eat into recovery. On those days, add a rep or two of reserve rather than forcing failure. A tool like the daily readiness check-in in the MBM app can help here: it scores your sleep, soreness and stress and suggests when to start at a lighter load, such as 85%.
How do I get the effort right every week?
- Log your reps and load. If you beat last week, the effort was there.
- Watch rep speed. When a rep slows to about half the speed of the first rep, you are near 1-2 RIR.
- Keep form fixed. If range of motion shrinks to finish the set, the set is over. The MBM app's Move IQ uses your phone camera to count reps and flag reps that miss range.
- Push harder late in a training block, then ease off for a lighter week.
Common questions
Is it bad to train to failure every set?
It is not dangerous on safe exercises, but it is inefficient. Fatigue builds fast, later sets suffer and recovery takes longer. Most lifters do better saving failure for the last set.
Do beginners need to train close to failure?
Beginners grow from moderate effort at first. Still, learning to push sets to 1-3 RIR within the first few months helps, since most new lifters stop far too early.
What is technical failure?
Technical failure is the point where you can no longer do a rep with the same form and range of motion. It often comes a rep or two before you physically cannot move the weight, and it is a sensible stopping point on big lifts.
Does training to failure help strength?
Not much. Strength seems to depend more on lifting heavy loads and practising the lift than on how close to failure you get.
Sources
Ready to apply this to your own program?
MBM builds 1:1 coaching programs around the exact principles above - weekly check-ins, weekly adjustments, and a plan that fits your life.