Do you need to train to failure?

What 'failure' actually means

The word 'failure' gets thrown around loosely in gym culture. Before you can decide whether you need it, you need to know what it actually is.

• Momentary muscular failure (MMF). The point where you genuinely can't complete another full rep with acceptable technique — the muscle has temporarily run out of force. This is true failure.
• Technical failure. The point where your form breaks down enough that continuing would shift load away from the target muscle or create injury risk. Most lifters hit this before MMF.
• Volitional failure. You choose to stop because it feels brutal — even though one or two more reps might technically be possible. This is how most 'going to failure' actually plays out.
• Concentric failure. Unable to complete the lifting phase of the rep. Eccentric failure means unable to control the lowering — rarely relevant to normal training.

Research on training intensity is talking about momentary muscular failure. When a study says participants 'trained to failure', it means they performed reps until they couldn't complete another — not until it felt uncomfortable.

Knowing where your sets actually end is the starting point for any honest evaluation of failure training. The RPE/RIR guide covers how to calibrate your effort ratings accurately.

Reps in reserve: a better way to measure effort

Reps in reserve (RIR) describes how close to failure a set was — specifically, how many more reps you could have done before hitting momentary muscular failure.

• RIR 0 — You couldn't have done one more rep. True failure.
• RIR 1 — One rep left in the tank. Very close to failure.
• RIR 2 — Two reps left. Hard, but clearly not failure.
• RIR 3 — Three reps left. Moderately challenging.
• RIR 4+ — Too easy to drive meaningful adaptation for most trained lifters.

Eric Helms, one of the researchers who helped establish RIR as a practical training tool, describes it as a way to standardise effort across sessions, exercises, and athletes. The advantage over percentage-based loading (%1RM) is that it accounts for daily variation in readiness — when you're undertapered or sleep-deprived, the same weight feels harder and leaves fewer reps in reserve.

Brad Schoenfeld, whose meta-analyses on hypertrophy are among the most cited in the field, has consistently noted that training must be taken close to failure to maximise motor unit recruitment — but 'close to failure' and 'to failure' aren't the same thing. RIR 1–2 already recruits the highest-threshold motor units that RIR 0 does. The marginal gain from that last rep is small; the fatigue cost isn't.

For a full breakdown of the RIR scale and how to apply it when programming rep ranges, see the RPE/RIR guide and the rep ranges article.

What the research says about proximity to failure

The central question: does training to failure produce more hypertrophy than stopping a few reps short? The answer is no — at least not meaningfully, when volume is equated.

A 2021 systematic review by Grgic, Schoenfeld, and colleagues examined studies comparing failure and non-failure training on muscle growth. Their conclusion: when total volume (sets × reps × load) is matched, training to failure doesn't produce superior hypertrophy compared with stopping 1–3 reps short. The growth stimulus is driven primarily by the high-effort reps near the end of a set — and those reps occur whether or not you complete the very last possible repetition.

Dr. Mike Israetel, co-founder of Renaissance Periodization, frames it through the 'effective reps' hypothesis: the final five or so reps of a set approaching failure are the primary growth stimulus. Reps performed with substantial reserve (RIR 5+) contribute little. Reps at RIR 1–3 provide most of the growth stimulus. The one rep at RIR 0 adds a small increment of stimulus while adding a disproportionate amount of fatigue.

Much of the early research on failure training was done with lower total volume. When you're doing 3–4 sets per muscle group per week, going to failure on every set may be necessary to generate enough total stimulus. At higher volumes — 10–20 sets per muscle group per week, where most intermediate lifters sit — each individual set doesn't need to be taken to the absolute limit, because the volume itself provides sufficient stimulus.

See the rep ranges article for how set and rep selection interact with proximity to failure in a full periodisation plan.

The fatigue cost nobody talks about

The failure-vs-non-failure debate tends to focus on stimulus — whether failure produces more hypertrophy. The cost side gets far less attention, which is exactly where most lifters feel the real difference.

Training to failure on compound exercises — squats, deadlifts, bench press, rows — generates a disproportionate amount of central and peripheral fatigue relative to the marginal additional stimulus it provides. That fatigue has real consequences.

• Impaired performance on subsequent sets. Go to failure on your first set of bench press and your second set will produce fewer reps at the same weight. The net effect on total session volume — sets × reps × load — is often negative compared with stopping 2 reps short across all sets.
• Impaired performance in subsequent sessions. Failure training on Monday means Wednesday's session starts from a lower baseline. This is why Israetel's MEV/MAV framework explicitly recommends leaving reps in reserve during the accumulation phase of a mesocycle — you need to recover well enough to train again.
• Injury risk on technical exercises. On movements where technique degrades under fatigue — deadlifts, overhead press, split squats — pushing to failure means the last 1–2 reps are performed with compromised form. The stimulus from those reps shifts away from the target muscle toward compensatory structures.
• Psychological fatigue. Sets taken to true failure are genuinely unpleasant. Over a 6–8 week mesocycle, demanding failure on every set creates a dread response that erodes training adherence — arguably the most important variable in long-term progress.

Israetel's Minimum Effective Volume (MEV) and Maximum Adaptive Volume (MAV) framework offers a useful lens: the goal of any given session isn't to maximise effort on each set, but to accumulate enough stimulus to force adaptation while staying below the recovery threshold. Failure training pushes recovery cost up without proportionally increasing stimulus.

If training feels unsustainable — soreness lasting 4–5 days, performance declining session to session — the deload guide explains when and how to reduce training stress so adaptation can actually happen.

When going to failure does make sense

The case against failure as a default isn't a case against failure entirely. There are specific contexts where taking sets to true failure is appropriate and productive.

• Isolation exercises on the last set. Cable curls, lateral raises, leg extensions, tricep pushdowns — movements where technical failure is unlikely and fatigue stays local. Taking the final set of a curl to failure carries almost no injury risk and minimal systemic fatigue.
• The last set of an exercise. If you're doing three sets of Romanian deadlifts, the first two might stop at RIR 2, and the third can go to RIR 0–1. You're not compromising any subsequent sets of the same exercise.
• Low-fatigue machines. Chest press machines, pec deck, leg press — exercises where the movement is guided and fatigue is contained. Going to failure on a pec deck is very different from going to failure on a barbell bench press.
• Testing your RIR calibration. Occasionally going to true failure teaches you what RIR 0 actually feels like, which makes your RIR 1–2 estimates more accurate on future sets. If you never reach failure, your sense of where your limit is gradually drifts.
• Intensification phases. Some programming deliberately includes failure work in the final week or two of a mesocycle — a peak intensity period before a deload. This is intentional, time-limited, and followed by recovery.

A practical rep-by-rep recommendation

Putting this into practice means building an RIR target into every set — not as a rigid rule, but as a framework that keeps your training in the productive effort range.

• 1Use RIR 2–3 on the first set of a compound lift. It feels challenging but leaves enough recovery capacity for the subsequent sets to also be productive. The first set is a primer, not where maximum stimulus comes from.
• 2Use RIR 1–2 on middle sets. As fatigue accumulates through the session, RIR 1–2 will feel progressively harder. The same RIR target produces a harder absolute effort as the session goes on.
• 3Allow RIR 0–1 on the final set of isolation exercises. This is where the cost-to-benefit ratio favours going close to or at failure — the exercise is safe to push, and there are no more sets of that exercise to compromise.
• 4Log your RIR honestly after each set. If you repeatedly log RIR 2 but could actually have done 4 more reps, you're training too conservatively. If you log RIR 2 but can't complete the same reps next session, your RIR was lower than you thought.
• 5Let the progression system adjust load. Consistently hitting RIR 3–4 means you're leaving too much in reserve and the weight should go up. Hitting RIR 0 without completing target reps means recovery is the bottleneck — not effort. See the progressive overload guide for how auto-regulation handles this automatically.

SammaGym uses RIR as the core input to its auto-regulated progression engine. After each set you rate how many reps were left in reserve, and after the session the app calculates whether to add reps, add weight, or hold steady at the same load. You don't need to manually decide whether you were close enough to failure — the system reads your RIR data and adjusts. Start logging at app.sammagym.com to see how your effort ratings track across a full mesocycle.