There is a scorecard that runs in the background of every gym session. It does not live in any app. Nobody taught it. But it is absolute: a set that ends with the bar stuck in your hands counts. A set that ends with two reps still possible does not. The scorecard has one rule, and the rule is failure.
The cost is quiet. Not the grinding itself, which at least feels productive, but what the grinding prevents. The fourth set that was supposed to happen but could not because the third set emptied the tank. The session cut short because accumulated fatigue turned the last exercise into a formality. The weeks where soreness lasted long enough to push a training day into the next one. The scorecard collects its tax on every session, and the tax is paid in work that never happens.
Do You Have to Train to Failure Every Set to Build Muscle?
A meta-analysis pooled the results of fifteen controlled trials and delivered a verdict the scorecard cannot argue with: training to failure and stopping a few reps short produced the same muscle growth. The difference was so small that no measurement in any gym could detect it. For strength, the gap was equally invisible.
Training to failure every set is not required for muscle growth. A meta-analysis of controlled trials found no measurable difference in muscle size or strength between training to failure and stopping one to three reps short. Stopping short preserves recovery capacity, allowing more total training volume per session, which is the stronger predictor of growth.
— Grgic et al. 2022 · Journal of Science and Medicine in Sport · 15 studies, ~400 participants
That is not a coaching opinion dressed as evidence. That is pooled data from hundreds of participants across studies that varied in exercise selection, body region, and design. The answer held regardless of how the question was sliced.
The mechanism underneath makes the verdict sharper. When every set runs to failure, the body needs longer to recover between sets and between sessions. Total work drops. And total work across a session is the variable that actually predicts how much muscle grows. Failure does not just fail to add a growth signal. It actively reduces how many productive sets the session can hold. The currency the scorecard was collecting, intensity per set, was competing against the currency that mattered more: volume across the session.
to failure
reps short
BLAMED: Failure intensity per set — grinding every rep to the point the bar stops
ACTUAL: Total volume across the session — productive sets the body can recover from and repeat
Stopping one to three reps short of failure keeps the growth signal intact while preserving the capacity to do more total work. The sets feel less dramatic. The session holds more of them. The math favors the lifter who racks the bar with something left.
For experienced lifters, the picture has one honest wrinkle. Two studies within the meta-analysis found a small hypertrophy edge for training to failure in people who had been lifting for years. The edge was real but narrow, and it came from only two trials. If you have been training seriously for a long time and your progress has stalled, occasional failure sets on your final set of an exercise might offer something. Treating that as a mandate for every set of every session is reading a footnote as the headline.
The scorecard was never measuring the right thing. A set that ends with two reps still possible delivered the same growth signal as the set that ended with the bar pinned to your chest. The difference was that the first set left enough in the tank for the next one to matter.
Failure is not the enemy. Treating it as a requirement is. And when lighter weights enter the equation, the rules shift entirely, because failure becomes the one condition that makes low-load training work at all.