You log “80 kg × 8”. But could you have completed one more rep, or five? Load and reps do not answer that question. RIR and RPE add this context: how much you think you had left at the end of a set.
A precisely logged RIR
is not necessarily an accurate RIR.
These tools are useful if you understand their central limitation: you are estimating a capacity you have not yet tested. Here is how to read them, why estimates can go wrong, and how to use them to track your training.
RIR: how many reps were left in the set?
RIR stands for repetitions in reserve. An RIR of 2 means: “I estimate that I could have completed two more full reps with this load and the same technique.” At 0 RIR, no additional rep seems possible.1
Illustrative example: 80 kg × 8 at 2 RIR means eight completed reps and approximately two more available, without extra rest or assistance. It does not mean the ninth or tenth rep was actually attempted.
Define what you mean by failure. Momentary muscular failure means being unable to complete the lifting phase despite maximal effort. Stopping because technique breaks down is a different endpoint. Changing your range of motion, using momentum or extending the pauses between reps also changes the task you are comparing.
RPE: which scale are we using in strength training?
RPE stands for rating of perceived exertion. Several scales exist: a general difficulty rating is not automatically an estimate of RIR. Here, we mean the set RPE scale anchored to repetitions in reserve, examined by Zourdos and colleagues.1
| RPE | RIR | Meaning |
|---|---|---|
| 10 | 0 | No more reps |
| 9 | 1 | One more rep |
| 8 | 2 | Two more reps |
| 7 | 3 | Three more reps |
| 6 | ≈ 4+ | Several reps left; a less precise anchor |
Near failure, RPE ≈ 10 − RIR is a useful shorthand. It expresses a scale convention, not an objective measurement. Do not mechanically extend the subtraction to very easy sets, where the anchors are less precise. An 8/10 cardio difficulty rating or a whole-session RPE cannot be converted this way.
Why it matters: the same set can hide different levels of effort
Eight reps with five left and eight reps with none left look identical in a log that only records “load × reps”. Yet you stopped at different points relative to your capacity that day. RIR helps describe that difference and adjust a training prescription.
For hypertrophy, Robinson et al.’s exploratory meta-regressions (2024) associate finishing closer to failure with greater muscle growth. The relationship with strength gains is much less clear. However, the researchers estimated RIR from study protocols: their findings establish neither a magic threshold nor an individual optimum.2
The 2026 ACSM position stand, synthesizing 137 systematic reviews, found no consistent effect of training to momentary muscle fatigue on outcomes. It also highlights heavier loads for strength.3 These analyses address different questions: training closer to failure and having to reach failure on every set are different prescriptions.
The practical decision therefore depends on your goal, the exercise and the overall program. Accumulating sets at 0 RIR is not a goal in itself. Conversely, believing you are close to failure when you have plenty left may put you further from the intended prescription.
RIR is not a measurement. It is an estimate.
More precisely, the RIR you report is a subjective estimate of your distance from failure. Load can be read from the plates and completed reps can be counted; remaining reps are anticipated. Entering a number in an app does not remove that uncertainty.
Imagine calling 2 RIR after your eighth rep, then continuing to failure without a break and completing five more. Your reserve at that moment was five reps under those conditions: you underestimated it by three. Conversely, calling 3 RIR but only completing one more rep means you overestimated your reserve.
These errors have different consequences. The first may make you stop earlier than intended; the second may take you to failure when you wanted to avoid it. Muscle burning, breathlessness or wanting to stop do not, by themselves, answer “how many more full reps can I complete?”
Do lifters actually get it wrong?
Halperin et al.’s exploratory meta-analysis (online in 2021; 2022 journal issue) reports an average underprediction of 0.95 reps (95% CI: 0.17–1.73), with substantial heterogeneity. That is an average directional bias, not an individual absolute error or a correction to automatically add.4
Opposite errors can cancel each other out in an average. A group result cannot tell you whether your last 2-RIR rating was accurate. “People can never estimate RIR” would therefore be just as misleading as assuming every estimate is correct.
A useful counterexample: in 24 trained people bench pressing at 75% of their maximum, Refalo et al. (2024) found an average absolute error of 0.65 reps across pooled 1- and 3-RIR predictions. Good accuracy is possible in that protocol; it is not guaranteed for every set or exercise.5
Load and proximity to failure affect estimation difficulty
Halperin’s synthesis suggests better accuracy when fewer total reps are possible, including with heavier loads, and closer to failure. These are tendencies, not guarantees.4 Relative load and remaining reps are different: a heavy weight can feel difficult from the first rep without failure being imminent.
In Mansfield et al. (2020), 20 trained men improved their accuracy across successive bench press and prone row sets. Knowing the load did not improve their estimates.6 This does not prove that any specific load always produces accurate ratings: exercise selection, set progression and the timing of the estimate matter.
The practical implication is straightforward: distinguishing “one or two more reps” is often more useful than claiming to know the exact difference between 7 and 8 RIR. Experience should not become an appeal to authority. Check your anchors on the exercises and rep ranges you actually use.
Why an imperfect tool is still useful
Bastos et al.’s scoping review (2024) describes using RIR to select or adjust loads around a target reserve, while acknowledging limits in accuracy.7 Its practical value is making otherwise hidden information explicit: how close the set seemed to your limit.
A prescription of “8 reps with about 2 RIR” can guide adjustments to your capacity that day. If the set seems much easier than intended, you can reconsider the load; if it ends at failure, you know the planned reserve was not maintained. The number informs a decision without making it on its own.
How to improve your RIR estimates in practice
These are practical guidelines, not a scientifically validated protocol that guarantees perfect accuracy.
- Define a reference rep. Keep range of motion, technique and rhythm consistent. Partial or assisted reps should not quietly become your new standard.
- Estimate immediately after the set. Ask how many additional reps were possible without extending the rest between them. Separate that answer from your general fatigue.
- Accept uncertainty. “About 2” may be more honest than a falsely precise decimal. If you are unsure, remember that uncertainty when comparing sessions.
- Occasionally calibrate when conditions allow. On a familiar exercise where stopping is safe, call your RIR and continue without a break to count the additional reps you can complete with the same execution. Use appropriate safety equipment or supervision; do not improvise a failure test under an unprotected barbell.
- Do not make this mandatory on every set. Calibration compares a prediction with an outcome. Stop for pain or loss of control; that is no longer a comparable test of muscular reserve.
- Learn exercise by exercise. An accurate curl estimate does not guarantee the same accuracy on a squat or a much longer set. Record recurring discrepancies and gradually adjust your anchors.
Connecting RIR to progress
Compare sufficiently similar situations: the same exercise and range of motion, with similar rest and exercise order. The following examples are illustrative.
80 kg × 8 — RIR ≈ 2
80 kg × 10 — RIR ≈ 2
More reps with a similar estimated reserve is a signal of improved performance. The same load and reps with reserve increasing across several sessions may also be consistent with progress.
However, moving from 2 to 3 RIR once does not prove improvement. You may have recovered better or simply estimated differently. Improved performance also does not directly measure muscle growth.
To put these signals into your training history, read how to know if you are progressing in the gym. For the other metrics, see how to analyze a strength training session (French article).
How Grunty puts RIR into context
Grunty keeps your set information — loads, reps and RIR/RPE ratings when provided — and puts it into your training history. The value comes from comparing what you completed and the reserve you reported across sessions.
Your RIR remains an estimate. Grunty cannot verify reps you never attempted or turn that number into an absolute truth. A repeated trend, alongside comparable loads and reps, provides more context than an isolated rating.
Voice logging lets you provide that information with fewer manual steps. Workout analysis then places it within the available context.
Scientific sources
Evidence syntheses describe broad trends; experimental studies show what happened under specific conditions. An exploratory review cannot establish one universally optimal RIR.
- 1. Zourdos MC et al. — 2016
Novel Resistance Training-Specific Rating of Perceived Exertion Scale Measuring Repetitions in Reserve.
Journal of Strength and Conditioning Research, 30(1), 267–275.
DOI : 10.1519/JSC.0000000000001049 · PubMed : 26049792 ↗ - 2. Robinson ZP et al. — 2024
Exploring the Dose–Response Relationship Between Estimated Resistance Training Proximity to Failure, Strength Gain, and Muscle Hypertrophy: A Series of Meta-Regressions.
Sports Medicine, 54, 2209–2231.
DOI : 10.1007/s40279-024-02069-2 · PubMed : 38970765 ↗ - 3. Currier BS et al. — 2026
American College of Sports Medicine Position Stand. Resistance Training Prescription for Muscle Function, Hypertrophy, and Physical Performance in Healthy Adults: An Overview of Reviews.
Medicine & Science in Sports & Exercise, 58(4), 851–872.
DOI : 10.1249/MSS.0000000000003897 · PubMed : 41843416 ↗ - 4. Halperin I et al. — 2022 (online publication : 2021)
Accuracy in Predicting Repetitions to Task Failure in Resistance Exercise: A Scoping Review and Exploratory Meta-analysis.
Sports Medicine, 52(2), 377–390.
DOI : 10.1007/s40279-021-01559-x · PubMed : 34542869 ↗ - 5. Refalo MC et al. — 2024 (online publication : 2023)
Accuracy of Intraset Repetitions-in-Reserve Predictions During the Bench Press Exercise in Resistance-Trained Male and Female Subjects.
Journal of Strength and Conditioning Research, 38(3), e78–e85.
DOI : 10.1519/JSC.0000000000004653 · PubMed : 37967832 ↗ - 6. Mansfield SK et al. — 2020 (online publication : 2020)
Estimating Repetitions in Reserve for Resistance Exercise: An Analysis of Factors Which Impact on Prediction Accuracy.
Journal of Strength and Conditioning Research.
DOI : 10.1519/JSC.0000000000003779 · PubMed : 32881842 ↗ - 7. Bastos V, Machado S, Teixeira DS — 2024
Feasibility and Usefulness of Repetitions-In-Reserve Scales for Selecting Exercise Intensity: A Scoping Review.
Perceptual and Motor Skills, 131(3), 940–970.
DOI : 10.1177/00315125241241785 · PubMed : 38563729 ↗
Bibliographic note: the Hughes et al. (2020) study on four exercises is excluded because it was retracted in 2021 following data-collection errors that invalidated its data. Retraction notice.