Dr. Pak on RPE and Reps in Reserve: How Accurate Do Lifters Really Need to Be?
UPLIFT IRON CLUBRPE and reps in reserve sound wonderfully precise. Call a set RPE 8, and you are saying roughly two good repetitions were left. Call it RPE 9, and you are saying one remained. That makes autoregulation feel almost mathematical: rate the set, adjust the load, and keep training at the intended effort even when sleep, stress or readiness changes.
But how accurate do lifters actually need to be? Dr. Pak Androulakis-Korakakis has spent years working in strength research and coaching, and he has also coauthored research examining how well people predict repetitions to failure. The practical answer is reassuring: RIR does not need to function like a laboratory measurement to be useful.
Across the available evidence, lifters are often wrong by roughly a repetition on average. Accuracy tends to improve when sets are closer to failure and when rep counts are not extremely high. That means the useful skill is not identifying the exact difference between 7 RIR and 8 RIR. It is reliably distinguishing an easy set from a moderate one, a hard set from a near-maximal one, and knowing when your estimate is drifting enough to change the training effect.
What RPE and RIR Are Actually Measuring
In lifting, the common RPE scale is tied closely to reps in reserve. An RPE 10 means no additional clean repetition was available. RPE 9 is about one rep in reserve, RPE 8 about two, and RPE 7 about three. Below that, the distinction becomes less important because the set is already relatively far from failure.
That last point is one reason lifters sometimes make RPE harder than it needs to be. If a warm-up feels like it could have been seven more reps or nine more reps, the practical decision is usually the same: it was easy. The rating becomes more useful as the set approaches the zone where effort affects programming decisions.
Research on RIR prediction supports that idea. A scoping review and exploratory meta-analysis that included Dr. Pak found that people tended to underpredict the repetitions they could still perform by about one rep on average. In other words, someone calling a set 2 RIR might frequently have been capable of roughly three more reps. Prediction also tended to improve closer to failure and on lower-repetition sets.
That error is not automatically a problem. If your program says to perform most hypertrophy work around 1–3 RIR, being off by a single repetition usually still leaves you in a productive neighborhood. The bigger issue is a systematic bias—calling sets RPE 8 when they are actually RPE 6 week after week, or turning every supposed RPE 8 into a grinder that was really RPE 10.
Why Accuracy Gets Better Near Failure
A hard set gives you more information than an easy one. Bar speed slows. Repetitions feel more demanding. Technique may begin to require more concentration. Breathing changes. Those signals make it easier to judge whether one, two or three clean reps remain.
Far from failure, those clues are weaker. The first five repetitions of a weight you could lift fifteen times may all feel nearly identical, so asking whether you have eight or ten reps left is mostly false precision. That does not make RIR useless; it just means the scale should be treated as a practical effort zone rather than a perfectly calibrated instrument.
Higher-repetition sets create another challenge. The discomfort from breathing, local burning and cardiovascular strain can make a set feel close to failure even when the target muscle still has several repetitions available. This is one reason a 5-rep squat at 2 RIR may be easier to judge than a 20-rep leg press at 2 RIR.
Exercise familiarity matters in practice as well. If you have performed the same bench-press setup for years, you know what your final successful repetitions usually look and feel like. A brand-new cable variation provides fewer anchors. That is why consistent exercise selection can help not only progression but also effort calibration.
A 2025 study in resistance-trained lifters found that RIR estimation was already reasonably accurate at baseline and could improve further during a training period that deliberately used changing RIR targets. That is useful because it suggests the skill can be practiced. You do not need to be born with perfect internal effort perception.
How Accurate Do You Need to Be in Real Training?
For most lifters, the best answer is accurate enough to make the right programming decision. If the plan calls for a top set at RPE 8, you want to stop before a grind while still using a load heavy enough to challenge you. If the plan calls for hypertrophy work around 1–3 RIR, you want the set hard enough to stimulate growth without accidentally turning every set into failure training.
Think in zones. An RPE 6–7 set should feel clearly submaximal. RPE 8 should be hard but controlled. RPE 9 should leave only one credible clean rep. RPE 10 is the end of the road for that repetition target. If you can consistently place sets in the right zone, being off by half a point or one repetition is rarely a disaster.
The easiest way to improve is occasional calibration. On an exercise where failure is reasonably safe—such as a machine press, leg extension or cable movement—predict your RIR and, occasionally, continue the set to technical failure. If you called 2 RIR and completed five extra clean reps, your internal scale needs adjustment. You do not need to do this often, and you do not need to test it on risky heavy compounds.
Video can help too. Watch how your final repetitions actually move. Lifters are sometimes surprised by how fast an alleged RPE 9 rep looked, or by how much technique changed before they admitted a set was maximal. Recording a few key sets gives you an external reference to compare with the feeling you had under the bar.
Most importantly, use your own ratings consistently. Two lifters may experience the same objective effort slightly differently. Autoregulation works because the rating helps guide decisions within your training, not because everyone in the gym must agree that your set was exactly 8.0.
The takeaway: Dr. Pak's work on RIR accuracy supports using the scale without worshipping the decimal point. Most lifters are close enough for RPE and RIR to guide training, especially near failure and on moderate-rep sets. Treat the numbers as useful effort zones, calibrate occasionally, and worry more about systematic misrating than being one rep wrong on a single Tuesday.
Powerlifting designs for lifters who track the bar, the load and the effort—not just the pump.
EXPLORE COLLECTION
