Training to Failure: When It Helps and When It Just Costs You
Failure training produces disproportionate fatigue for a modest hypertrophy return, and for a hybrid athlete that trade is usually bad. Where it earns its place, where reps in reserve is the better tool, and how to judge proximity to failure honestly.
- Most of the growth stimulus is available in the last few reps before failure, without reaching it.
- Failure disproportionately increases fatigue and time needed to recover.
- Never on compound barbell lifts. Technique fails before the muscle does.
- Reasonable on machines, cables and isolation work, occasionally.
- Most people are worse at judging proximity to failure than they think — usually stopping too early.
There is a persistent belief that a set only counts if it ends with the bar stopping. It produces satisfying training and, over a block, generally less progress.
The relevant question is not whether failure works. It is whether the last rep is worth what it costs — and for most people, on most exercises, it is not.
What the last reps do
The useful mechanism first, because it explains everything else.
As a set progresses and fatigue accumulates, your nervous system recruits progressively larger motor units to keep producing force. The final few reps before failure are where the highest-threshold motor units are recruited and where mechanical tension on the working fibres is highest.
That is where the growth signal lives. Which is genuinely why "train hard" is good advice — a set stopped six reps early does very little.
But the signal is largely present before failure is reached. Research comparing sets taken to failure against sets stopped one to three reps short generally finds similar hypertrophy, with failure occasionally edging ahead and often not.
Meanwhile the fatigue cost of that last rep is not proportional. It is considerably higher.
The actual trade
| Stopping 2 reps short | Going to failure | |
|---|---|---|
| Growth stimulus | Nearly all of it | All of it |
| Fatigue generated | Moderate | High |
| Effect on later sets | Small | Substantial |
| Recovery time | Normal | Extended |
| Technique risk | Low | Elevated |
The key row is effect on later sets. A first set taken to failure reduces what you can do in sets two and three. Since total quality volume across the session and the week is a primary driver of adaptation, maximising one set frequently costs you the session.
For a hybrid athlete this compounds, because the extra fatigue is drawn from the same recovery budget as your running. A Monday leg session taken to failure does not just cost Monday — it shows up in Wednesday's intervals.
Reps in reserve
The practical tool is RIR: how many more reps you could have done.
- RIR 0 — failure
- RIR 1 — one left
- RIR 2 — two left
- RIR 3 — three left
Useful defaults:
- Hypertrophy work: RIR 1–3
- Strength work on compounds: RIR 2–4
- Anything technical or heavy: RIR 3+
- The last set of an isolation exercise: RIR 0–1 is fine occasionally
RIR is a better prescription than a percentage of your one-rep max, because it self-adjusts. On a bad day, 85% might be RIR 0; on a good day it might be RIR 3. Prescribing effort rather than load handles that automatically.
The honest problem with RIR
Most people are bad at estimating it, and the error goes in a specific direction.
Untrained and intermediate lifters consistently stop earlier than they think. Someone reporting RIR 1 frequently has three or four reps left. This is why "train closer to failure" is common and correct coaching advice for beginners — not because failure is required, but because their RIR 2 is genuinely RIR 5.
Calibrating it takes practice. The most efficient way is to occasionally take an isolation exercise to true failure — a leg extension, a cable curl, a machine press — purely to recalibrate what the last rep actually feels like. Do that once every few weeks, on something safe, and your estimates on the important lifts get considerably better.
The tell for a genuine RIR 1: the final completed rep moves noticeably slower than the ones before it. If the bar speed is unchanged, you were not close.
Where failure is genuinely fine
Isolation work. Curls, lateral raises, leg extensions, calf raises, cable work. Low technical demand, low systemic cost, easy to bail. If you want to take something to failure, take these.
Machines. Same reasoning, plus you cannot be trapped under them.
The final set of an accessory. A reasonable place to spend it.
Bodyweight work. Push-ups and bodyweight rows are safe to fail on.
Where it is a bad idea
Heavy compound barbell lifts. Squats, deadlifts, bench, overhead press. Two reasons:
- Technique degrades before the muscle does. The rep that fails is preceded by a rep with a rounded back or a collapsed knee, and that is where injuries happen
- The systemic fatigue is enormous. A squat set to failure affects everything for days
On deadlifts specifically, never. The lift has no safe failure mode and there is no version of a grinding final deadlift rep that is worth its risk.
Anything before a running session. Obvious once stated, routinely ignored.
In a calorie deficit. Recovery capacity is already reduced. Failure training in a deficit accumulates fatigue faster than you can clear it — the context is in the calorie deficit that doesn't kill your training.
During a race build. Every unit of recovery is spoken for.
What this means for a hybrid athlete
The calculation is more one-sided than for a pure lifter.
A bodybuilder taking sets to failure is spending recovery on the only thing they are training. You are spending it on one of two, and the endurance half will notice.
So the default should be RIR 2 on almost everything, with failure reserved for occasional isolation work when you want to recalibrate. This is not a soft approach — the sets are still genuinely hard, and stopping two reps short of failure at a heavy load is not comfortable. It simply stops short of the point where the cost curve turns vertical.
The same logic drives the loading recommendations in the deadlift for endurance athletes and the only six lifts a runner needs: heavy, brief, and stopped before the wheels come off.
The summary
Train hard enough that the last few reps are genuinely difficult and moving slowly. Stop one to three reps before the bar does.
Take isolation work to failure occasionally, to keep your judgement calibrated. Never take a heavy compound there.
You will leave most sessions feeling you could have done more. Over a block, you will have done more — because you were able to come back on Wednesday.
Questions people ask
Should you train to failure?
Rarely, and selectively. Most of the hypertrophy stimulus is present in the final few reps before failure, so training one to three reps short captures nearly all the benefit at substantially lower fatigue cost. For a hybrid athlete whose recovery is shared with running, that trade strongly favours stopping short.
What is reps in reserve?
A way of describing how many more reps you could have completed at the end of a set. RIR 2 means you stopped with two reps left in the tank. It is a more practical prescription than a fixed percentage of your one-rep max, because it adjusts automatically for how you feel on the day.
Does training to failure build more muscle?
Slightly, at best, and inconsistently in the research. What it does reliably is generate considerably more fatigue, which reduces the volume you can perform in subsequent sets and sessions. Since total quality volume is a primary driver of hypertrophy, chasing failure often reduces growth over a training block rather than increasing it.
How close to failure should you train?
For hypertrophy, one to three reps in reserve on most sets. For maximal strength, two to four — heavy compounds are trained for force production and technique, not for exhaustion. Anything taken past that should be a deliberate exception on a safe exercise, not a default.
Is failure training bad for recovery?
It increases both the depth and duration of recovery needed, and the effect is disproportionate — the last rep costs far more in fatigue than it delivers in stimulus. For someone also running or racing, that additional fatigue is drawn from the same budget as their endurance training, which is why it so often shows up as a worse session two days later.
Keep reading
Bulgarian Split Squats: Why They Hurt So Much and Why You Should Do Them
The most hated accessory in strength training is also one of the highest-return exercises for anyone who runs. What makes them disproportionately brutal, how to set up so you train the right thing, and how to program them without wrecking your week.
Core Training for Runners: What Actually Transfers
Most 'weak core' diagnoses are weak hips, and most core work is chosen for burn rather than function. What the trunk actually does during running, the four exercises worth keeping, and why endless crunches change nothing.
Grip Strength: The Silent Limiter in Every Hybrid Race
Your farmers carry does not end because your legs quit. Grip is the most trainable and most neglected quality in hybrid sport — and it correlates with a surprising amount beyond the gym.