Training to Failure vs Not: Is It Better for Muscle Growth?

Training to failure isn't bad, but for muscle growth it doesn't beat ending your sets one or two reps short of failure. What it does do is tire you out more and slow down recovery, so it pays to be selective: isolation work, machines and the last set, far more than a heavy squat.

If you're not yet sure what RIR (reps in reserve) means, start with our guide on what RIR is and how to use it. This article picks up where that one leaves off: what happens when RIR hits zero.

What training to failure means

"Failure" doesn't always mean the same thing, and that explains some of the confusion in the research:

• Momentary muscular failure. You can't complete another rep however hard you try. It's RIR 0 in the strict sense, and it's the definition the more careful studies use.
• Technical failure. A common coaching term rather than a scientific category: you stop when the next rep wouldn't come out with good form. It arrives before muscular failure, often a rep earlier.

The Refalo et al. (2022) meta-analysis ran into exactly this problem: many studies called different things "failure". So it separated momentary muscular failure from any other definition of "set failure".

What the research says about failure and hypertrophy

There are several meta-analyses and recent trials. Read together, they paint a fairly consistent picture.

Meta-analyses find no clear advantage for failure.

• Refalo et al. (2022), 15 studies: pooling every definition of failure, there was a "trivial" advantage for failure. But looking only at momentary muscular failure, the advantage disappeared. The authors conclude there's no evidence that failure is superior and point to a non-linear relationship: getting closer to failure doesn't always add more.
• Grgic et al. (2022), also 15 studies: no significant differences in either strength or hypertrophy. One interesting detail: in the subgroup of already-trained people, failure did show a small hypertrophy advantage. Their conclusion: failure doesn't seem necessary, but it doesn't seem harmful either.
• Vieira et al. (2021), 13 studies: failure produced more hypertrophy overall, but the difference vanished when both groups did the same volume. In other words, part of the advantage came from the failure group simply doing more work.

But proximity to failure does matter. Robinson et al. (2024) reanalysed the literature by estimating the RIR in each study and found that muscle grows more the closer to failure sets end. The authors themselves urge caution: the analysis is exploratory and each study's RIR is estimated. Taken together, the data say that being close matters, not hitting zero on every set.

Head-to-head trials point the same way.

• In Refalo et al. (2024), trained lifters worked one leg to failure and the other at 1–2 RIR for eight weeks on the leg press and leg extension. Quadriceps thickness increased almost identically in both legs.
• In Hermann et al. (2025), using single-set routines, the group that went to failure tended to gain a little more on several hypertrophy measures than the group leaving 2 RIR, although the differences were modest. When volume is minimal, pushing that one set may add a bit more.

Strength: failure matters even less here

For maximal strength, failure doesn't seem to add anything:

• Robinson et al. (2024) found a negligible relationship between RIR and strength gains: people gained similarly across a wide range of effort.
• Grgic et al. (2022) and Vieira et al. (2021) agree that when volume isn't matched, strength improves slightly more when you don't go to failure. Vieira found the same for power.
• Pareja-Blanco et al. (2017) compared stopping squat sets at 20% or 40% velocity loss. The 20% group gained the same strength, improved their vertical jump more and did 40% fewer reps. The 40% group gained somewhat more muscle in parts of the quadriceps.

The cost of failure: fatigue and recovery

This is where the evidence is most consistent:

• More immediate fatigue. A meta-analysis of 20 studies (Vieira et al., 2022) found that training to failure causes a bigger drop in performance, a larger metabolic response, more muscle damage and higher perceived exertion than stopping short.
• Longer recovery. Morán-Navarro et al. (2017) compared three sets of 10 reps to failure with doing only 5 of the 10 possible reps. After failure, recovery was clearly slower between 24 and 48 hours, even against a protocol with the same total volume.
• A near-linear relationship. In Refalo et al. (2023), six sets of bench press to failure cut bar speed by 25%, versus 13% at 1 RIR and 8% at 3 RIR. At 24 hours, failure and 1 RIR were still down, and by 48 hours the differences had gone. On top of that, the closer to failure, the worse people felt and the more soreness they reported.

In practice: failure can leave you with less in the tank for your next sets and your next session, without a bigger growth stimulus to show for it.

When training to failure makes sense, and when it doesn't

This table is common coaching practice built on the data above, not a proven rule:

Situation	Failure?	Why
Isolation (curls, extensions, lateral raises)	Yes, especially on the last set	Little overall fatigue; failure is easy to control
Machines and cables	Yes, in moderation	The load is guided and stopping is simple
Heavy compounds (squat, deadlift, bench press)	Better at RIR 1–3	More fatigue and slower recovery; form degrades under heavy load
Beginners on technical lifts	No	Technique and learning to estimate RIR come first
One set per exercise	Can be worth it	In Hermann et al. (2025) failure added a little at minimal volume
Maximal strength goal	Not needed	Strength gains are similar at higher RIR
Deload week	No	The goal is recovery

Two more details. In Hermann et al. (2025), participants estimated their RIR more accurately on the bench press than on the squat: on big lifts it's easier to misjudge how close you are to failure. And a set to failure counts as one set, not two: if total volume is your concern, check how many sets per muscle per week you're doing and where you sit relative to your MEV, MAV and MRV.

How to use failure without overdoing it

• Save failure for the last set of an exercise. Keep the earlier ones at 1–2 RIR so you don't sink the performance of the rest.
• Use it to calibrate. An occasional set to failure on a machine shows you how far off your RIR estimates are. If you'd rather think in terms of perceived exertion, here's the difference between RIR and RPE.
• If you always end up at failure, check the load. With double progression, hitting failure before the bottom of your rep range means the weight is too heavy. If you reach the top of the range but only at failure, it's usually wiser to consolidate before adding weight.

How MUTARA handles it

In MUTARA every exercise in your routine has a target RIR (RIR 1–2 by default), and each set is logged with its weight, reps and RIR. If you finish a compound exercise at failure for four sessions in a row, the app warns you that leaving one or two reps in reserve pays off more over time. Isolation exercises don't get that warning: there, failure is a reasonable choice.