Training to Failure or Using RIR? What the Evidence Shows
Training to failure is not required to build muscle. Learn how to use RIR, get sufficiently close to your limit and manage fatigue.
There’s a line I heard so many times in the gym that it became almost religious:
“The only rep that counts is the one you didn’t think you could do.”
I believed it. For years, every set of mine ended with the bar stalling halfway up, someone yelling “it’s all you!”, and me getting off the bench convinced I’d done my job.
Then I started training in a way closer to powerlifting: heavy sets, clean technique, stopping with two or three reps still in the tank. No grinding. No shouting.
Two things happened, and one of them I didn’t expect.
The first: I got stronger. Not marginally — clearly, and faster than when I was grinding.
The second was the one that made the rest make sense. During the period when I took everything to failure, the thing consistently going up wasn’t the load. It was my stress. Worse sleep, worse mood, that sensation of permanently running on a 40% battery. I had a ready-made and wrong explanation for it — I called it “intensity” and felt proud of it.
It wasn’t intensity. It was debt.
Let’s get to what the evidence says, because my experience here isn’t the exception — it’s basically the standard finding.
First, the vocabulary
Three terms you’ll see constantly, worth being precise about:
RIR (Repetitions In Reserve) — how many reps you could still perform with good technique when you decide to end the set. RIR 0 is failure. RIR 2 means “I had two more in there.”
RPE (Rating of Perceived Exertion), on Zourdos’s resistance-training-adapted scale, is the same thing seen from the other side: RPE 10 = RIR 0, RPE 9 = RIR 1, RPE 8 = RIR 2, and so on.
Failure has more than one definition, and confusing them generates a lot of pointless argument:
- Technical failure: you can still move the load, but only by breaking form. This is where the set should end, almost always.
- Concentric failure: the bar stops mid-ascent and won’t move, even with form dissolved. This is what most people mean by “failure.”
When this article says “training to failure,” it means the concentric kind. And when I talk about stopping short, I mean stopping short of even technical failure.
What the evidence shows
This is one of the most studied questions of the last decade, and the reviews converge quite a bit.
The systematic review with meta-analysis by Refalo and colleagues (2023) found that reaching momentary muscular failure was not superior to stopping short. That answers an important question — do you have to fail? — but it does not mean every distance from failure produces the same stimulus.
A more recent meta-regression by Robinson and colleagues (2024) treated RIR as a continuous variable. The result was more nuanced: hypertrophy tended to increase as sets ended closer to failure, although the exact relationship remains uncertain; for strength, RIR had little relationship with the outcome. A 2024 trial also found similar hypertrophy between failure and 1–2 RIR, with less fatigue in the group that stopped short.
The meta-analysis by Grgic and colleagues (2021) looked at strength and hypertrophy together and found something even more interesting: for hypertrophy, failure and non-failure came out equivalent; for strength, results tended to favor not going to failure.
The honest summary is: failure does not appear necessary, but getting sufficiently close matters for hypertrophy. For many people, 1–3 RIR offers a useful compromise between stimulus and fatigue. And that is where the part that changes everything comes in — because tying on outcome does not mean tying on cost.
The cost nobody accounts for
The study by Morán-Navarro and colleagues (2017) is, in my opinion, the most clarifying of all. It tracked the recovery time course after sessions taken to failure versus sessions stopped short.
The finding: sessions to failure produced fatigue markers — loss of neuromuscular performance, metabolic and hormonal alterations — that persisted substantially longer. On some markers, full recovery took 24 to 48 hours more.
That reorganizes the whole equation. If going to failure:
- does not deliver more hypertrophy,
- may deliver less strength,
- and costs an extra day or two of recovery…
…then it isn’t a choice about intensity. It’s a high-interest loan, repaid out of the volume of your next sessions.
And “your next sessions” are where muscle is actually built. You do not grow because of one heroic set on Tuesday. You grow from weeks of sets performed sufficiently close to failure — and one needlessly heroic Tuesday set can rob quality from Thursday.
This is what the training literature calls the stimulus-to-fatigue ratio: what matters is not how much fatigue you generated, but how much stimulus you got per unit of fatigue. Failure can worsen that ratio, especially on heavy compounds and when volume is high. The last rep can be stimulating — but it is also usually the slowest, most expensive rep and the one most likely to degrade technique.
ILLUSTRATIVE EXAMPLE: TWO WEEKS, SAME EXERCISE
──────────────────────────────────────────────────
STRATEGY A — everything to failure
Session 1: 4×8 @ 80kg (ground out)
Session 2: 4×7 @ 80kg (still fatigued)
Session 3: 3×7 @ 80kg (volume dropped)
Session 4: 4×8 @ 80kg (back to square one)
→ 15 productive sets, load stagnant
STRATEGY B — RIR 2
Session 1: 4×8 @ 77.5kg
Session 2: 4×8 @ 80kg
Session 3: 4×8 @ 82.5kg
Session 4: 4×8 @ 82.5kg + 1 rep
→ 16 productive sets, load climbing
──────────────────────────────────────────────────
The strategy that "goes easier" progressed more.That isn’t a pretty theoretical chart. It’s basically what happened to me.
The problem: your perception is wrong
Now the caveat that stops this article from becoming an excuse — and it’s a serious one.
There’s a consistent body of research showing that people are bad at judging how many reps they have left. And the error has a direction: most people underestimate their own capacity. When they say “I can’t do one more,” there were frequently three, four, sometimes five reps still available.
The error is largest among:
- beginners, who haven’t yet learned what being near the limit actually feels like;
- heavy compound lifts, where systemic discomfort arrives before local muscular failure;
- people who’ve never failed at anything, because they’ve never calibrated the ruler against reality.
And here’s the beautiful paradox: to train at an honest RIR 2, you need to know what RIR 0 is. If you’ve never been close, your “RIR 2” is probably RIR 5 — and then you aren’t being strategic, you’re being easy on yourself with a scientific justification.
Calibration is a skill, and it trains. The practical way to acquire it:
Once every few weeks, on the last set of an isolated, safe exercise (curl, leg extension, lateral raise, a machine), go to genuine concentric failure. Not for the stimulus — for the reference.
Notice what happens on the two preceding reps: bar speed collapses, the gap between reps grows, form starts to give. That deceleration is your RIR marker. It’s far more reliable than your feeling of “I’m tired.”
Concentric speed is a useful practical indicator without equipment, but it is not an exact RIR converter: the relationship varies by exercise, load and person. When a rep clearly slows down, you are probably approaching failure. Learning your own pattern makes the estimate less arbitrary.
Where failure still earns its place
None of this means failure is banned. It means it’s an expensive tool, and expensive tools get used where the return justifies them:
Worth it:
- On the last set of an exercise, not every set.
- On isolation work and machines: curls, triceps, calves, leg extensions, flyes. Low technical risk, low systemic cost, and where failure feedback is clearest.
- On high-rep sets (15+), where the absolute load is low.
- When you need to calibrate your RIR ruler.
- In specific phases with little load available — training at home, traveling, no equipment — where going near failure is the only way to generate enough stimulus.
Not worth it:
- On squats, deadlifts, heavy bench and the like. Recovery cost is maximal, form degrades where it’s most dangerous, and the return is the same.
- On every set, of anything.
- On the first set of an exercise — you’ve just wrecked the next three.
- When you’re short on sleep, high on stress, or already deep in a hard week. That’s exactly when the recovery math doesn’t close.
PRACTICAL RIR GUIDE
──────────────────────────────────────────────────────
Heavy compounds (squat, deadlift, bench) RIR 2–3
Accessory compounds (rows, presses) RIR 1–2
Isolation / machines RIR 0–1
Last set of an exercise 1 lower
Deload week RIR 4–5
Beginner, first 6 months RIR 2–3, focus on form
──────────────────────────────────────────────────────The part that makes RIR actually work
There’s an obvious risk in everything I’ve written, and it would be dishonest not to name it: “RIR 2” is the perfect excuse to train light and feel smart about it.
Stopping short of failure only works if it’s coupled to one thing: verifiable progressive overload.
Meaning there has to be external proof that load, reps or volume are climbing across weeks. Without that, you don’t have a strategy — you have a feeling. And feeling, as we just saw, is precisely the faculty the research shows to be miscalibrated.
In practice, the pairing that works is this:
Stop 2 reps short of failure — and log every set.
If the number in your log climbs month over month, your RIR is honest. If it’s been flat for eight weeks, your “RIR 2” became RIR 5 without you noticing, and the problem isn’t the strategy — it’s your execution of it.
This is exactly why logging matters more for people who train this way than for people who grind everything: those who go to failure get the brutal feedback of failing on the bar; those who stop short need data in place of that feedback. Logging exercise, load and reps — in D-Fit or anywhere you’ll actually look afterwards — is what turns “I think I’m progressing” into a line that either rises or doesn’t. Without that line, RIR is faith.
What I’d tell the guy who ground everything out
- The last rep isn’t magic. It can be stimulating, but it also carries more fatigue and is not mandatory.
- Stop 1 to 3 reps short on most of your work, especially heavy compounds.
- Save failure for isolation and last sets — and even then, sparingly.
- Calibrate your ruler by going to failure occasionally on something safe, just to remember what it feels like.
- Learn to read bar speed. It’s more honest than your sensation.
- Log everything, because stopping short is only a strategy if the load is climbing.
- If your stress is rising faster than your bench, the variable to change isn’t your diet. It’s your proximity to failure.
Point 7 is mine. It took me years to see, because I was convinced that permanent exhaustion was the price of taking it seriously. It wasn’t. It was just a Tuesday session charging interest on Thursday.
Bottom line: training to failure is not mandatory, but hypertrophy still requires sufficiently challenging sets. For most work, especially compound lifts, ending with 1–3 reps in the tank usually balances stimulus, technique and recovery. Prove that the load or repetitions are increasing, and let failure be an occasional tool rather than an identity.
References:
- Refalo MC, Helms ER, Trexler ET, Hamilton DL, Fyfe JJ. “Influence of resistance training proximity-to-failure on skeletal muscle hypertrophy: a systematic review with meta-analysis.” Sports Medicine. 2023.
- Grgic J, Schoenfeld BJ, Orazem J, Sabol F. “Effects of resistance training performed to repetition failure or non-failure on muscular strength and hypertrophy: a systematic review and meta-analysis.” Journal of Sport and Health Science. 2021.
- Morán-Navarro R, Pérez CE, Mora-Rodríguez R, et al. “Time course of recovery following resistance training leading or not to failure.” European Journal of Applied Physiology. 2017.
- Zourdos MC, Klemp A, Dolan C, et al. “Novel resistance training-specific rating of perceived exertion scale measuring repetitions in reserve.” Journal of Strength and Conditioning Research. 2016.
- Steele J, Endres A, Fisher J, Gentil P, Giessing J. “Ability to predict repetitions to momentary failure is not perfectly accurate, though improves with resistance training experience.” PeerJ. 2017.
- Sánchez-Medina L, González-Badillo JJ. “Velocity loss as an indicator of neuromuscular fatigue during resistance training.” Medicine & Science in Sports & Exercise. 2011.
- Davies T, Orr R, Halaki M, Hackett D. “Effect of training leading to repetition failure on muscular strength: a systematic review and meta-analysis.” Sports Medicine. 2016.
- Robinson ZP, Pelland JC, Remmert JF, et al. “Exploring the dose-response relationship between estimated resistance training proximity to failure, strength gain, and muscle hypertrophy: a series of meta-regressions.” Sports Medicine. 2024.
- Refalo MC, Helms ER, Robinson ZP, et al. “Similar muscle hypertrophy following eight weeks of resistance training to momentary muscular failure or with repetitions-in-reserve in resistance-trained individuals.” Journal of Sports Sciences. 2024.
