Training to failure: do you need it to build muscle and how close to get, according to science

Published on 27 September 2026

Reaching failure on every set doesn't build more muscle than stopping one or two reps short, but it does bring more fatigue. What four meta-analyses say, how close to failure to go on each exercise and how to measure RIR without fooling yourself.

In a study published in 2024, a team of researchers trained 18 people with gym experience for eight weeks. Each participant trained one leg to muscular failure on every set and the other leaving one or two reps in the tank. At the end, they measured quadriceps thickness: 0.181 cm of growth with failure and 0.182 cm without it. The same hypertrophy, with considerably less fatigue in the leg that didn't go all the way.

That result neatly sums up what science knows today about training to muscular failure: you don't need it to build muscle or strength, but getting close to it does matter. Here's exactly what failure is, what the meta-analyses say, how close to get on each exercise and how to measure it without wrecking yourself on every set.

What muscular failure is (and what it isn't)

Momentary muscular failure is the point at which, despite trying as hard as you can, you can't complete another rep with acceptable technique. It isn't when the set starts to hurt or when the bar slows down: it's when the rep doesn't go up.

The opposite is finishing the set with reps in reserve (RIR): the ones you could still have done. A set at 2 RIR is a set where you stop knowing you had two reps left. The scale became popular after the work of Zourdos and colleagues (2016), who linked it to the rating of perceived exertion (RPE) scale: an RPE of 10 equals 0 RIR, an RPE of 9 equals 1 RIR, and so on.

Is training to failure necessary to build muscle? What the meta-analyses say

At least three meta-analyses directly compare training to failure with not doing so, and all three point in the same direction.

• Grgic and colleagues (2022) pooled 15 studies in young adults. They found no significant differences in either strength or hypertrophy between going to failure and stopping close to it. Only in the subgroup of trained people did a small hypertrophy advantage for failure appear.
• Vieira and colleagues (2021), with 13 studies, saw more hypertrophy with failure in the overall analysis, but the difference disappeared when both groups did the same volume of work. For strength and power, not going to failure gave equal or better results.
• Refalo and colleagues (2023), also with 15 studies, found a "trivial" advantage for failure when all definitions were pooled, and no difference when true muscular failure was compared with not reaching it.

The piece that completes the picture comes from Robinson and colleagues (2024). Instead of comparing "failure or no failure", they estimated how many reps in reserve each programme left and looked for a continuous relationship. The result: strength gains were similar across a wide range of RIR, while hypertrophy increased as sets ended closer to failure. The authors urge caution because the analysis is exploratory, but the practical message fits with everything above: to grow, you need to get close; going all the way adds little.

The cost of failure: fatigue and recovery

If failure doesn't give you more muscle, what's the problem with always chasing it? Fatigue. In the study by Morán-Navarro and colleagues (2017), ten trained men did the bench press and squat with three protocols: 3 sets of 10 reps to failure, 6 sets of 5 with the same load (the same total volume, but far from failure) and 3 sets of 5. Going to failure caused a bigger acute drop in jump performance and bar velocity, and delayed recovery by up to 24-48 hours. The non-failure protocols recovered clearly sooner.

Velocity-based training shows it from another angle. Pareja-Blanco and colleagues (2017) trained 22 young men on the squat for eight weeks: one group cut each set when velocity dropped by 20% and the other when it dropped by 40% (much closer to failure). They gained similar strength, but the 20% group improved their jump more (9.5% versus 3.5%) while doing 40% fewer reps. The 40% group achieved slightly more hypertrophy in two quadriceps muscles, but lost some of their fastest fibres.

In other words: failure comes at a price. It leaves you worse off for the next session, lowers the quality of the following sets and, on heavy exercises, increases the risk of your technique breaking down just when you're handling the most load. We explain this in more detail in how to manage training fatigue.

How close to failure to go on each exercise

Not every exercise costs the same when you push it to the limit. A biceps curl to failure barely generates any general fatigue and carries no risk; a heavy squat to failure leaves you beaten up for a couple of days and can end with the bar on top of you. Here's a reasonable guide based on the above:

That's the same logic as in Refalo's 2024 study: leg press at 2 RIR and leg extension at 1 RIR, with the same hypertrophy as going to failure on both. If you train with light loads or at home, getting close to failure becomes more important, because with little weight the first reps barely stimulate the muscle; we cover it in how to build muscle at home.

The problem with RIR: almost everyone stops short

Training "at 2 RIR" sounds precise, but it requires knowing how many reps you have left, and that's hard. The meta-analysis by Halperin and colleagues (2022), with 12 studies and 414 participants, found that people underestimate by almost one rep on average (0.95) how many they have left before failure, with a lot of variation between studies. Accuracy improves when the prediction is made near the end of the set and in sets of 12 reps or fewer. Surprisingly, training experience didn't improve it.

Put simply: when you think you have two left, you probably have three. Three simple ways to calibrate yourself:

• Do a set to failure every now and then on safe exercises (machines, isolation). That way you learn what your limit with that load really feels like.
• Write down your reps: if you do the same number on your last set as on your first, you were further from failure than you thought. A progressive overload log makes it obvious.
• Pay attention to speed: when a rep becomes clearly slower than the previous ones without your technique changing, you're usually 1-3 reps from failure.

Training to failure as a beginner

When you're starting out, your body responds to almost any stimulus and your technique isn't stable yet. Going to failure on squats or deadlifts in the first few months has little to gain and quite a lot to lose. The practical recommendation is to work at 2-4 RIR on the basic lifts and use failure only on machines or isolation work, more as a learning tool than as a method. We go into detail in the beginner's routine.

As you gain experience, your sets need to get closer to the limit to remain a sufficient stimulus, and that's when it does make sense to finish isolation work at 0-1 RIR and push the odd final set.

How to apply it to your routine

• By default, end your sets 1-3 reps from failure. That's the zone where almost all the stimulus is concentrated, for a fraction of the fatigue.
• Save failure for the last set of isolation or machine exercises. Reaching failure on the first set sinks your performance on the ones that follow.
• Never go to failure on heavy compound lifts without safety measures: there, the risk far outweighs the benefit.
• If you increase volume, stay a bit further from failure; if you do few sets, get closer. The two variables offset each other, as we explain in volume versus intensity.
• In deload weeks, go up to 4-5 RIR to recover without losing what you've gained.

The Progrevia routine generator already includes the target effort for each exercise (RIR) based on your level and the type of movement, so you don't have to decide on every set how hard to push.

Progrevia · Updated 27 September 2026 · How we write our guides

Informational content based on the available scientific evidence: it does not replace advice from a healthcare professional. If you have an injury or a medical condition, or are pregnant, check with one before changing your training or diet.