Training to Failure: When It Helps and When It Hurts

What training to failure actually means

First, let me define what we are talking about because people use "failure" loosely. True muscular failure means you cannot complete another rep with acceptable form no matter how hard you try. The muscle literally cannot generate enough force to move the weight through the range of motion.

This is different from:

• •Stopping because it hurts. Pain is not failure. Burning is not failure.
• •Form breakdown. If your form falls apart but you could grind out another ugly rep, that is technical failure, which is different from true muscular failure.
• •Giving up mentally. A lot of people stop 3-4 reps short of failure and think they went to failure because the set felt hard.

Most people have never actually trained to true failure. They think they have, but they have not. A set taken to real failure on something like leg press or leg extensions is one of the most unpleasant experiences in the gym. It is your muscles screaming, your brain telling you to stop, and you pushing through both of those signals until the weight physically does not move.

The question is not whether training to failure is effective. It is. The question is whether it is more effective than stopping a few reps short, and whether the cost (recovery, injury risk, fatigue) is worth the benefit.

The case for training to failure

There are legitimate reasons to train to failure, and the bro-science crowd is not entirely wrong about its benefits.

Maximum motor unit recruitment. As a set progresses and muscle fibers fatigue, your body recruits progressively larger motor units to keep moving the weight. By the time you reach failure, essentially all available motor units in the target muscle are firing. Henneman's size principle (1957) established this hierarchy, and subsequent research has confirmed that taking sets closer to failure increases motor unit recruitment.

Hypertrophy stimulus. Mechanical tension is the primary driver of muscle growth, and the reps closest to failure produce the most mechanical tension per rep. Those last 3-5 reps of a hard set (often called "effective reps" or "stimulating reps") are where the majority of the growth stimulus comes from. Burd et al. (2012) demonstrated that low-load training to failure produced similar hypertrophy to high-load training to failure, supporting the idea that proximity to failure matters for growth.

It teaches you what hard actually feels like. This is underrated. If you have never experienced true failure, you have no calibration for what "2 reps from failure" or "RPE 8" actually feels like. Training to failure occasionally recalibrates your effort gauge so your submaximal training is actually at the intensity you think it is.

Accountability. When the prescription is "go to failure," there is no guesswork. You push until you cannot. For people who tend to sandbag their training, failure sets ensure they are actually working hard enough.

The case against training to failure

And here is why you should not take every set to failure:

Recovery cost is massive. Training to failure dramatically increases muscle damage, neural fatigue, and recovery time. Morán-Navarro et al. (2017) compared sets taken to failure versus sets stopped 2-3 reps short. The failure group took significantly longer to recover (48+ hours versus 24 hours) and reported more soreness. If you go to failure on every set, your total weekly volume will suffer because you cannot train as frequently.

Injury risk increases. When you are grinding out that last rep, your form deteriorates. Joints are in compromised positions. Muscles are fatigued and less able to stabilize. This is when strains, tweaks, and tears happen. A torn pec on the last rep of a bench set taken to failure. A pulled hamstring on that last deadlift rep. I have seen it plenty of times.

Diminishing returns. The difference in muscle growth between a set taken to 1 rep from failure (RIR 1) and a set taken to true failure (RIR 0) is very small. But the difference in recovery cost is large. You get maybe 5% more stimulus for 40% more fatigue. The math does not work out.

It is unsustainable. Nobody can train to failure on every set of every exercise, every session, for months on end without burning out, getting injured, or regressing. The most productive training programs use failure sparingly, not as a default.

What the research actually says

This is where it gets interesting, because the research does not fully support either camp.

A meta-analysis by Grgic et al. (2022) examined studies comparing training to failure vs not training to failure and found: there was no significant difference in muscle hypertrophy between groups that trained to failure and groups that stopped 1-3 reps short of failure. The hypertrophy outcomes were essentially the same as long as the non-failure group was training close to failure (within about 3 reps).

However, the failure group experienced more fatigue, longer recovery times, and in some studies, higher rates of overtraining symptoms. So you get the same muscle growth but at a higher cost.

Santanielo et al. (2020) found similar results: stopping 2-3 reps short of failure produced comparable hypertrophy to going all the way to failure, with less fatigue accumulation.

For strength, the picture is slightly different. Izquierdo et al. (2006) found that training to failure reduced explosive strength and power output compared to stopping short. This makes sense because maximal grinding sets train slow-twitch endurance, which is the opposite of explosive power.

The current evidence suggests that the sweet spot for most training is stopping 1-3 reps short of failure (RPE 7-9 or RIR 1-3). Close enough to failure to get the growth stimulus, far enough away to recover and train frequently.

RIR and RPE: a better way to think about intensity

Instead of the binary "failure or not failure" debate, modern programming uses rating scales to quantify how close to failure each set is.

RIR (Reps in Reserve): How many more reps could you have done?

• •RIR 0 = true failure, no more reps possible
• •RIR 1 = could have done 1 more rep
• •RIR 2 = could have done 2 more reps
• •RIR 3 = could have done 3 more reps

RPE (Rate of Perceived Exertion): On a 1-10 scale, how hard was the set?

• •RPE 10 = true failure
• •RPE 9 = could have done 1 more (RIR 1)
• •RPE 8 = could have done 2 more (RIR 2)
• •RPE 7 = could have done 3 more (RIR 3)

These scales are imperfect because self-assessment is tricky (especially for beginners who tend to overestimate how close to failure they are). But they provide a useful framework for programming intensity.

For hypertrophy, most of your working sets should fall between RPE 7-9 (RIR 1-3). For strength, RPE 8-9.5 on your primary lifts. Save RPE 10 (true failure) for specific situations.

When to go to failure

Training to failure is a tool, not a religion. Here are the scenarios where I actually recommend it:

Last set of an isolation exercise. If you are doing 3 sets of lateral raises, take the last set to failure. The injury risk on a lateral raise is minimal, the recovery cost is low because it is a small muscle, and you get a strong growth stimulus to finish off the session. This is the safest and most productive place to use failure.

Machine exercises. Machines lock you into a safe path, which means form breakdown at failure is less risky. Leg press, chest press, leg extensions, hamstring curls. These are all fair game for failure sets because the machine prevents you from getting into a dangerous position.

Cable exercises. Similar logic to machines. The cable provides a consistent line of pull and there is minimal injury risk at failure.

Calibration sets. Once every few weeks, take a set to true failure on your main lifts just to recalibrate your RPE/RIR perception. This ensures your "RPE 8" is actually an 8 and not a 6 because you have lost touch with what hard feels like.

The last week of a training block. If you are running a periodized program with a deload coming up, pushing to failure in the final week before deloading can be effective. You accumulate maximum stimulus, then recover during the deload. This is how most good hypertrophy programs work.

When to stay away from failure

Compound barbell exercises with heavy loads. Heavy squats, deadlifts, bench press, and overhead press taken to grinding failure are where injuries happen. The risk-to-reward ratio is terrible. A failed squat rep with 315 on your back is a completely different animal than a failed set of cable curls. Stop 1-2 reps short on these lifts.

Early in the session. If you go to failure on your first exercise, every subsequent exercise in that session will suffer. Your nervous system is fried, your stabilizers are fatigued, and your performance drops. Save failure for the end of the session, not the beginning.

High-frequency training. If you are squatting three times per week, you absolutely cannot go to failure every session. The accumulated fatigue will crush you by week two. High-frequency programs work because each session is submaximal (RPE 7-8), allowing recovery between sessions.

When you are in a caloric deficit. Dieting reduces your recovery capacity. Going to failure while cutting is a recipe for muscle loss and injury. During a cut, keep your intensity honest (RPE 7-8) but save failure for perhaps one set at the end of your session.

Beginner lifters. Beginners need to learn proper form and motor patterns. Going to failure means grinding through ugly reps, which reinforces bad movement patterns. Beginners should stop with 2-3 good reps left in the tank and focus on perfect form. Failure training can come later once the movement patterns are ingrained.

Different types of failure

Not all failure is the same, and understanding the differences helps you program more intelligently.

Concentric failure: You cannot complete the lifting (concentric) portion of the rep. The bar gets stuck halfway up on a bench press, or you cannot curl the dumbbell past the halfway point. This is what most people mean by "failure." It is a reasonable stopping point for most exercises.

Technical failure: Your form breaks down before you reach true muscular failure. Your squat turns into a good morning, your curl turns into a full-body swing, your row turns into a hip thrust. I think this is actually where most people should stop on compound exercises. When form breaks, the set is over.

Eccentric failure: You cannot even control the lowering (eccentric) phase. The weight drops. This only happens with extreme fatigue and is generally not something to train to intentionally. If the weight is dropping on you, you went too far.

Isometric failure: You cannot hold the weight in a static position. Think of a wall sit where your legs give out and you slide down the wall. This can happen at specific sticking points in compound lifts.

For practical purposes, stop compound exercises at technical failure and take isolation exercises to concentric failure. Going beyond concentric failure (forced reps, negatives) should be used very sparingly and only with a spotter.

How to program failure training intelligently

Here is a practical framework for incorporating failure into your program:

Compound movements (squat, bench, deadlift, overhead press, rows):

• •Sets 1-3: RPE 7-8 (RIR 2-3). Build up, practice form, accumulate volume.
• •Last set: RPE 9 (RIR 1). Push it, but keep 1 in the tank.
• •Technical failure at most. Never grind ugly reps on these.

Secondary compounds (dips, chin-ups, lunges, incline press):

• •Most sets: RPE 8-9 (RIR 1-2).
• •Last set: RPE 9-10. Failure is acceptable here because the risk is lower.

Isolation exercises (curls, lateral raises, extensions, flyes):

• •Most sets: RPE 8-9.
• •Last set: RPE 10 (true failure). Go for it. Low risk, good stimulus.
• •Consider a drop set or rest-pause on the final set for extra volume.

Sample session applying this framework:

In this example, you train hard throughout but only go to true failure on the last set of machine and isolation exercises. The compound exercises stay 1-2 reps from failure. Total failure sets: 3 out of 16 total sets. That is about 20% of your volume at failure, which is a reasonable ratio.

If you are currently going to failure on everything, try pulling back to this approach for 6-8 weeks. I would bet money that your strength goes up, your joints feel better, your sleep improves, and your muscle growth is at least the same if not better. Less is often more when it comes to training to failure. The guys who figure this out earlier tend to have longer and more productive lifting careers than the ones who redline every session until something breaks.