Recovery

How often should you train to failure?

Training to failure is one of the few topics where both extremes have loud, credible advocates. One camp says a set that is not taken to failure has left stimulus on the table. The other says failure buys a small amount of extra stimulus at a large cost in fatigue and injury risk.

Both are describing something real, and the argument is usually unresolvable because it is being had without the variable that determines the answer.

The variable is how many sets you are doing.

Failure and volume are one decision, not two

Take four hard sets a week on a muscle group. That is a modest amount of work, near the bottom of the useful range. Taking those four sets to failure is a defensible way to make sure that a small number of sets carries enough stimulus to be worth doing.

Now take twelve sets a week on the same muscle, all to failure. That is not the same decision made three times over. Fatigue does not scale politely with set count when every set is maximal, and the recovery demand of twelve failure sets is disproportionately worse than three times the demand of four.

So "should I train to failure" has no answer on its own. What has an answer is "should I train to failure at the volume I am currently doing", and the threshold worth using is around eight hard sets per week on a muscle group.

Above eight sets a week, all at genuine failure, is the specific combination worth changing. Below it, failure is a much more reasonable default.

What to do when you are in that combination

The change is not to remove sets. It is to leave one or two reps in reserve on most of them.

The order matters here, and it goes against most people's instinct. If you are stalled at ten sets a week of maximum-effort training, the instinct is to cut volume, because volume is the thing everyone talks about cutting. But backing off the effort is a smaller, cheaper change: you keep every set, you keep the frequency, you keep the exercise selection, and you shed a large fraction of the fatigue.

Stopping one rep short costs very little of the stimulus and a lot of the cost. That is not a close trade at high set counts, and it is reversible in a way that removing sets from your program is not.

Removing sets is the next lever, not the first one. It matters in the other direction too: cutting sets can drag a muscle group under the four-set-a-week floor, at which point you have swapped a fatigue problem for a stimulus problem.

The part where your zero is probably not zero

All of the above assumes you know what failure is. Most people, honestly, do not.

Reps in reserve is a self-report, and self-reports of proximity to failure are consistently optimistic. Lifters who believe they are at zero are frequently at two or three. This is not a character flaw; the sensation of a set being very hard arrives well before the point at which you actually cannot complete another rep.

Which creates a real problem for any rule that keys off effort. A logged zero that is actually a two means the rule fires on a lifter who is not in the situation the rule is about.

So a claimed zero is worth checking rather than believing. The practical way to build a reference is to occasionally take a set to genuine failure on something safe: an isolation movement, a machine, anything where the failing rep does not put you underneath a loaded barbell. Do it rarely, and use it to recalibrate what the last rep actually feels like.

If you have never done this, your RIR numbers are a scale with no zero marked on it.

Where this sits among the other fixes

If a lift is stalled, effort is not the first thing to look at, and reaching for it too early is a common way to waste a fortnight.

The order that works is structural first, then lifestyle, then training variables.

• Weekly volume comes first, because it is a single number and both of its failure modes have unambiguous fixes. Under four sets a week, add. Over ten, cut to ten immediately.
• Lifestyle second. Chronic short sleep, a deep deficit, sustained stress and recent illness all lower what you can recover from, and no adjustment to effort survives three weeks of four-hour nights.
• Effort third, and this is where the eight-set threshold applies.
• Rep range and exercise selection after that.

The value of a fixed order is not that each step is individually clever. It is that you stop picking whichever explanation appeals to you today, and you stop trying the expensive fixes before the cheap ones.

The short version

Failure is not the enemy and it is not the goal. It is a setting whose correct value depends on how much work you are doing.

Count your hard sets per muscle group per week. Under eight, going to failure is fine. Above eight and going to failure on all of them, that combination is doing more to your recovery than it is doing for your progress, and pulling back to one or two in reserve is the cheapest useful change available to you.

Common questions

Should every set go to failure?

It depends far more on how many sets you are doing than on failure itself. At a low set count, training to failure is a reasonable way to make sure the stimulus is there. Above roughly eight hard sets a week on one muscle group, every set to failure produces more fatigue than most people recover from.

Is leaving reps in reserve less effective?

Stopping one or two reps short costs surprisingly little stimulus and reduces fatigue a lot. That trade is usually worth making, and it becomes clearly worth making as your set count climbs.

How do I know if my zero RIR is really zero?

Most people overestimate how close they are. If you have never taken a set to genuine failure you probably do not have a reference point, and the honest way to build one is to occasionally take an isolation set past where you would normally stop.

What should I change first if I am stuck and training to failure?

Check your weekly set count first. If it is above ten, cut it to ten. If it is between eight and ten and every set is genuinely to failure, backing off to one or two reps in reserve is the cheaper change and should be tried before removing sets.

Does training to failure make a stall worse?

It can, because it raises fatigue without reliably raising stimulus. A stall while doing high volume at maximum effort is the specific case where doing less is more likely to help than doing more.