Somewhere around minute seventy-five, you stop lifting and start performing lifting. You know the moment. The bar is moving, the sets are getting logged, and you're technically still in the gym. But you're scrolling between sets, and the fifth set of the fourth exercise is a ritual rather than a stimulus. You tell yourself that more time means more progress. Then you drive home tired, and it isn't clear that any of the last half hour did anything.
Almost everyone who lifts has been told a number for this. Sixty minutes is the usual one. After that, the story goes, cortisol spikes and your muscle starts to break down. The story is tidy and widely repeated, and it's a poor way to decide when to stop.
Where the 60-minute rule comes from
Cortisol does rise during hard exercise. It's a normal part of mobilizing energy, and it rises and falls with training intensity, not with a stopwatch. Nothing in the research I'm aware of shows a cliff at the one-hour mark where lifting turns catabolic. The number survives because it's round, and because it points at something true: sessions can get too long.
The real problem is subtler than a hormone threshold. It's about what the work at minute ninety is actually worth.
What actually decays as a session drags on
Three things erode over a long session, and none of them is cortisol.
Your sets get worse. Fatigue accumulates. If you bench 185 for 8 on your first set and 185 for 5 on your fourth, the later sets didn't fail. They were still hard sets, and hard sets close to failure are what drive growth. But a set you can only grind through because the exercises before it drained the same muscles is a different kind of work than one you start fresh.
Your attention frays. Research on mental fatigue, including Samuele Marcora's 2009 work on cyclists, found that people who first did a demanding thinking task reported higher perceived effort and quit earlier on a physical task. Their muscles weren't weaker. The effort just felt harder. A long session isn't only a physical load. It's a sustained demand on focus, technique and decision-making, and those drift before your muscles do.
Your technique softens. When you're depleted, the rep you'd call clean at the start of the day gets sloppy. That lowers the quality of the stimulus and nudges up the injury risk.
What these share is that you can't see them on the clock. A well-run ninety minutes with long rests on heavy compound lifts can be fine. A distracted forty-five can be wasted. So the question isn't "how long?" It's "is this still working?"
The last-set test
Here's the idea worth taking from this article: judge the session by the quality of your most recent working set, not by elapsed time.
It's simple to run. Any given exercise has an expected rep count at a given weight and effort. Say you did 3 sets of squats at 225 and landed 8, 7, 7. That's a normal fatigue curve: small, predictable drops. Then you move to accessory work and notice something different. The same movement pattern suddenly drops by half between sets, your form changes shape, or you can't recall what you did on the previous set.
That's your signal. Several signs together mean the session has stopped paying for itself:
- Reps fall off a cliff from one set to the next at the same weight, beyond what your recent sessions show.
- You need much more rest than usual to get a similar set.
- Your technique is visibly different from your first set.
- You're adding sets mostly because you planned to, not because you're still pushing.
This is a more honest rule than any fixed duration. A strong, rested lifter doing three heavy lifts with five-minute rests may need ninety minutes and still be sharp at the end. Someone doing a quick full-body session with short rests may be spent in forty.
Why more is not automatically better
The usual instinct is that volume equals results, and up to a point it does. Weekly training volume is one of the better-supported drivers of muscle growth. But the same research tends to show diminishing returns: each additional set adds less than the one before. At some point the extra sets cost more in recovery than they give back.
There's also a plain practical cost. A session that leaves you wrecked affects tomorrow's sleep, appetite and the quality of the next workout. If the fourth hour of your week's training is paid for by a worse next session, you haven't gained volume. You've moved it around.
The flip side matters just as much. Shorter isn't automatically smarter either. If you finish in thirty minutes every time with plenty left in the tank, you're likely leaving stimulus on the table. The target is the span where your sets are still productive.
What to do if your session runs long
Most long sessions are long for boring reasons, not heroic ones. Rest periods stretch because you're on your phone. Warm-ups sprawl. Equipment waits eat twenty minutes. Before you cut any real work, trim the dead time, because the dead time is where the clock goes while the training stays the same.
If the work itself is too much, you have two clean fixes. Drop the lowest-value exercises, which are usually the third accessory movement for a muscle that's already been hit hard. Or split the volume across two sessions in the week rather than cramming it into one. Research comparing training frequency generally finds that when total weekly volume is matched, how you divide it matters less than how much you do and how hard you push. That means a session you can finish fresh is a perfectly good way to hit the same weekly total.
Your next moves
- Run the last-set test today. Pick your first main lift and note the reps on your first and last working sets. Write them down. A drop of one to three reps across sets at constant weight is normal. A collapse is a signal.
- Time one full session, start to finish, including warm-up. Don't change anything. Just learn your real number. Most people guess low by fifteen to twenty minutes.
- Find the dead time. Mark every stretch where you sat for more than your planned rest. Cut that first, before you cut a single set.
- Set a stop rule, not a stop time. Write one sentence before you train, like "I stop adding accessory sets when form changes or reps drop by more than half." Decide it when you're fresh, not when you're tired.
- Compare your next session to this one. If you stopped earlier and your top sets were just as good, your extra time wasn't buying anything. If your performance fell, you cut too deep. Adjust by one exercise and try again.
The number that matters is the one you wrote down
Notice what each of those steps depends on: comparing a set today with the same set last time. You can't run the last-set test from memory, because memory is generous. It remembers your best set and forgets that the fifth one fell apart. A log that's fast enough to use between sets, without breaking your focus, is what turns this from a vague feeling into a rule you can follow. That's what we built Rep for: quick entry, clean history, and your previous session right there when you need it. You pay once, and there's no subscription. If you want to see how your own sets decay across a session, you can look at it at rep.lumenlabs.works.