It is late afternoon on a Monday and you’ve just arrived at the gym for Day 1 of this week’s training after a tough day at work. You pull out your training log. Today’s workout prescription:
Back Squat: 125 kg x 4 reps x 4 sets.
125 kg is 2.5 kg heavier than last week’s squats, and you’re excited to be making incremental progress on this lift. You finish your last bit of pre-workout drink and get to work.
The first few warm-up sets feel routine, but as the weights climb you notice they aren’t moving as easily as you remember from the previous week. In fact, by the time you’ve reached your final warm-up set, you’re not even sure how you managed to finish the last rep. But the program says today’s load is 125 kg — so you load the bar, take a rest, then give it a go.
You perform your first rep (which feels heavy and slow), second rep (feels heavier and slower still), and third rep (feels very heavy and the bar almost grinds to a halt). Since the program says 4 reps, you go for your fourth and final rep, getting a third of the way back up only to ride it down to rest on the spotter pins. You crawl out from under the bar feeling frustrated and defeated. You’ve failed the last rep of your first programmed set. “Am I getting weaker? How am I supposed to do three more sets of this?” You skip to the next exercise in an effort to salvage your workout, feeling guilty that you have not completed the program for the day.
If this sounds familiar, you are not alone. Many of our lifters, particularly those coming off of linear progression-type programs, have lived this experience of “getting stuck failing.” Trainees will often report this experience alongside aches, pains, low motivation to continue training, and a general frustration with their actual results compared to their expectations. Typically they will have re-attempted the same strategy multiple times, under the assumption that the load must be forced upwards in order to make progress, only to come to a similar end each time.
The only situation where a given input leads to a consistent, predictable output is among machines. The reality is that humans are not machines, so the kind of performance variability described above is inevitable. Fortunately, there are ways of managing training intensity that account for these fluctuations in performance and can lead to a lifetime of more enjoyable, effective training.
Enter autoregulation.
What is Autoregulation?
Autoregulation describes a means of selecting and adjusting training intensity based on your performance during a given training session. Several methods can be used to autoregulate training intensity, including RPE (Rate of Perceived Exertion), RIR (Reps in Reserve), and VBT (Velocity-Based Training). It is important to note that autoregulation is not exclusive to resistance training; its use extends to aerobic and endurance exercise, where we regularly use autoregulatory metrics to prescribe training intensity as well.
To better understand the role of autoregulation in training, let us explore the concept of “load”. In the context of training, the external load placed upon a trainee refers to the absolute physical work performed — for example, lifting 150 kg for 8 repetitions, or running 5 km at a specific pace. Performing this external load elicits a unique psychological and physiologic response in the trainee. This response is termed the internal load, and it acts as the stimulus for the specific adaptations that result from the training session.
Take, for example, an endurance athlete who is programmed to perform a 5 km run in 25 minutes. This prescription represents the external load. During this particular effort, the athlete’s heart rate increases and they experience a particular perception of difficulty. The heart rate response and perception of effort are measures of the athlete’s internal load.
This is a nuanced and critically important point: the external load (here, the 5 km performance) is not itself the stimulus for adaptation; if it were, we would expect similar responses across the population, or within an individual in different contexts. Rather, the external load is merely a strategy for eliciting a psychological and physiological response, which is unique to each trainee. It is this unique response — the internal load — that ultimately drives the individual’s adaptation. Therefore, differences in internal load relate to the wide variation we see between individuals in their response to a particular training prescription.
The internal load can vary according to a number of factors other than the absolute external load. These include the athlete’s training status, psychological or emotional state, nutritional status, genetics, and environment, among many others. For example, consider the differences in heart rate response and perception of effort during the same 5 km, 25 min running effort when it is performed by a novice runner compared to an elite endurance runner. Or consider the experience when that same run is performed under significant negative psychological stress versus in a positive emotional state, when undernourished versus well-nourished, or in crisp, cool fall weather versus sweltering, humid summer heat. These same external loads would be expected to generate different internal load responses — and thus different adaptations — both between these athletes, as well as within each athlete in different situational contexts.
Again, humans are not machines. This is why, provided the same external load, the experience of internal load — and, consequently, training adaptations — will vary both between individual trainees, and within trainees in different contexts.
What is RPE (Rate of Perceived Exertion)?
Let’s turn our attention now to RPE, or Rate of Perceived Exertion. More recently adapted for use in resistance training (for more on the history of RPE in strength training, see our BBM Podcast with Mike Tuchscherer), the RPE scale ranges from 1 to 10. For those who are new to the use of RPE, a useful heuristic to apply at first is relating it to the estimated number of repetitions left before failure, also known as “Repetitions in Reserve” (RIR). In this context, an RPE 10 represents a maximal effort, where no additional repetitions could be completed at the same weight. A 9 RPE therefore describes a near-maximal effort, where one more repetition could be completed before failure; an 8 RPE denotes a difficult, albeit clearly sub-maximal effort, where two more repetitions could be completed at the same weight, and so on. See Figure 1 below.
Figure 1: Barbell Medicine Coach Tom Campitelli’s “Quick and Dirty Guide to Rate of Perceived Exertion” is a helpful illustration of the RPE scale. Note, this scale begins at 5 RPE, as it is rare to see prescriptions lower than this in your training program.
Intensity is often described using a percentage of an established 1-repetition maximum in resistance training programs. This approach quickly loses validity since normal performance variability means trainees are not typically capable of the same 1RM on any given day. Fortunately, training intensity can also be described using the RPE scale, which accounts for this problem since loads will change based on performance in real-time.
In the case of using RPE in strength training, the lifter must estimate the difficulty of their work set after completing it. Those lifters new to RPE or accustomed to percentage-based programming may at first find it difficult to decipher their training program. Traditional programs might pair reps and sets with an absolute load or a percentage of a one-repetition maximum, for example:
Back Squat 315 lbs x 5 reps x 3 sets
Back Squat 80% of 1RM x 3 reps x 3 sets
While simple, these prescriptions are rigid and do not allow for the normal variations in performance or readiness to train on a given day, which can lead to the failure and frustration described above. Instead, when using an RPE-based program, reps and sets are paired with a rating from the RPE scale. For example, on Day 1 of training, your program may read: “Back Squat 4 reps @ RPE 8 x 3 sets.” This means that you will perform three work sets of four repetitions each. Here, you will work up to a weight that you can perform for four reps, with an estimated two reps left in reserve for each set. The load at which this effort target is met will vary — heavier on days when you are performing well, and lower on days when performance is decreased.
In the context of aerobic training, metrics like RPE allow for prescriptions that account for differences in internal load. RPE-based prescriptions can also be delivered in practical terms for ease of understanding. Here, subjective markers for pace, such as “fairly easy pace that can be sustained,” and respiratory rate, like “can speak in short sentences but not sing,” can be used to gauge RPE. The RPE scale for aerobic training begins at 6 RPE, as it is rare to see prescriptions lower than this for aerobic work in your programming (see Figure 2).
Figure 2: The RPE scale for aerobic training.
A prescription using RPE for an aerobic training session might look like this: “Low-intensity steady-state (LISS) conditioning on a stationary bike or rower: 30 minutes @ RPE 6-7.” Here, a trainee will row or pedal at a relatively easy or conversational pace, and at an effort they estimate they can maintain for the prescribed duration of the exercise bout.
Trainees who are new to aerobic training may need to stop a few times throughout the session due to a lack of muscular endurance in the arms and back on the rower, or through the quadriceps on the bike. As aerobic fitness improves, trainees will cover longer distances on the rower and burn more calories on the bike within the same time frame at a similar level of effort. For example, on Week 1 you cover 8 miles within 30 minutes @ RPE 6 on the air bike, versus 9 miles on Week 8 for the same duration at a similar level of effort. Increases in performance at the same effort level are therefore one way to track progress for supplemental aerobic training.
What are the Pros and Cons of RPE?
To quickly recap, autoregulation methods like RPE accounts for variability in performance by allowing the lifter to adjust training loads based on real-time psychological and physiological feedback. This means that on any given day that you are training, weights may increase or decrease relative to a previous training session based on your performance on this particular day.
A common criticism of using RPE in training is that RPE is fundamentally subjective, and that the training load is therefore influenced by how the lifter “feels” on any given day. Some view this as an approach that gives lifters an excuse to avoid working hard in training. These views are problematic for a number of reasons. In fact, the subjective nature of RPE is precisely what makes it so useful. Since each individual’s unique psychological and physiological response to an external load drives their ultimate adaptation, it is wise to pay attention to these factors in training rather than deliberately ignore them through a myopic focus on performance numbers alone. RPE conveniently provides a composite metric of internal load with all of the relevant variables (such as training status, psychological/emotional state, nutritional status, genetics, environment, etc.) “baked into” the numerical rating.
Let us review an example: say a lifter using a linear progression-based program intends to work up to a near-maximal set of 5 reps on the deadlift. Their target weight is 107.5 kg. This is based on their 105 kg set from their deadlift session the prior week. (In an RPE-based program, this may look like “5 reps @ RPE 9.”) Here, the lifter works up in weight and finds they can only complete 3 of the 5 prescribed reps at 107.5 kg on this day at an RPE 10 effort. In effect, they cannot finish their prescribed workout, and the desired internal load — and thus the desired training effect — was not elicited. Using RPE, the lifter could have instead adjusted the load down (to, say, 97.5 kg for 5 reps, which on this day may elicit the desired RPE 9 effort), completing their volume prescription and receiving the intended training stress for this day without incurring excess fatigue.
Contrary to the common view that autoregulation inevitably leads to people avoiding hard work, RPE can be a useful tool to add more weight to the bar during particular training sessions or over longer training periods where performance is trending upwards. Let us consider another example: the following week, the same lifter on the same linear progression-based program is prescribed a lower target weight — say, 95 kg for 5 reps — based on their failed performance in the deadlift the prior week. On this day, the lifter finds that 95 kg performed for 5 reps actually feels easier than expected, that is, they estimate they could have done 3 — and possibly even more — reps before reaching failure. Using RPE, the lifter could have instead adjusted the load upwards (to 110 kg for 5 reps, for example, corresponding to RPE 9), incurring the intended internal load — and thus training adaptation — for this day. Here, again, a fixed progression misses the mark, as it would have delivered a lower external load than was necessary to elicit the desired effect for the day. In both the instances described, a more flexible approach allows the lifter to progress by imparting a more specific training stimulus within the context of their training program.
To summarize: RPE is subjective, but this does not mean that the lifter who uses RPE selects weights completely arbitrarily on any given day. The same lifter in question, having deadlifted 105 kg for a near-maximal set of 5 the previous week, does not expect in the following training session with the same protocol to suddenly deadlift 150 kg — a whopping 40 kg heavier than their most recent deadlift effort — or, by the same token, 40 kg less than they did the previous week. More likely, they have a relatively small range of target loads in mind. Paying attention to RPE can help calibrate working weights within this range based on the lifter’s performance during this particular training session. More importantly, “perfect accuracy” is not necessary for it to be a useful tool in training.
The advantage of using RPE over percentage-based prescriptions based on “historical” maxes is especially apparent at higher relative intensities. Experienced trainees regularly report variations in strength performance, and the ramifications of “overshooting” or “undershooting” in training outcomes can be substantial. Say, for example, that a coach wants to prescribe a 1-rep squat to mimic a first attempt in competition. In a percentage-based program this might be described as “a single @ 90% (based on a previous 1-rep max).” If the lifter’s current performance is high, this weight might be too light to be considered an opening attempt (undershooting), or, conversely, much too heavy if the lifter is experiencing a dip in performance for that day (overshooting). In both cases, a strict percentage prescription leads to loading that misses the goal for the day. This is because the training program according to external load is guided by a pre-recorded 1RM that does not reflect the lifter’s potential for performance on this particular day.
It’s easy to see how performance might vary over the long term, but many coaches fail to account for the fact that marked variations in performance can occur even within the same training phase. The data from last week was only accurate last week and determining working weights for the following training session based exclusively on this data can be counterproductive.
Lastly, using RPE may allow for greater autonomy and self-efficacy in a lifter compared to expecting a program or a coach to determine the external load. There is value in being able to accurately assess your potential under the bar and adjust your absolute load accordingly, both in training and in competition. You are ultimately possessed of information that others could not accurately know or apply to your efforts.
A few notes and Frequently Asked Questions regarding weight selection:
- On target weights: While there is nothing wrong with having progressively heavier target weights in mind, we caution against viewing this as the only way forward. Training programs written to force perfectly linear strength progress are often short-sighted and can set trainees up for disappointment when their performance inevitably varies or they experience a setback, which often leads to unnecessary panic about “wasting time.” Instead, we recommend accepting and embracing this unavoidable performance variability and meeting it with optionality, or the ability to course-correct, rather than bury our heads in the sand about how humans actually adapt over time.
- On linear progression: It is important to note that RPE and linear progression do not have to be at odds with one another — provided that linear progress happens without exceeding the prescribed training stress. That said, there’s no law requiring you to use RPE — but most people, wittingly or unwittingly, are making subjective considerations that account for internal load during their training.
- On RPE-based calculators and percentage charts: RPE percentage charts can provide lifters with estimates of target weights for particular rep and RPE pairings. The function of these charts is to ballpark weights only, and ideally, they should be used only when the lifter has recent training data on the lift in question. It must be said that the outcomes of these charts are seriously limited in terms of accuracy and generalizability across lifts, across people, and even across phases of an individual’s training.
Consider, for example, that a lifter at the end of a high-repetition training cycle uses the percentage chart to calculate their estimated 1RM based on a recent set of 8 reps @ RPE 9. When the lifter attempts a single @ RPE 8 during the new phase of training, they find that their estimated 1RM is lower than previously predicted by the percentage chart. This does not mean that the lifter hasn’t gotten stronger. They’ve just gotten stronger or more skilled in a different context or rep range (in this case, sets of 8). The data from that set of eight cannot be extrapolated to accurately inform weight selection for a very different rep range (singles).
No system is perfect and RPE has potential pitfalls. Newcomers to this method can underestimate or overestimate their level of exertion relative to their physical potential. A lifter who aggressively and consistently underrates their RPE is likely to incur so much fatigue that it stymies their longer-term progress. By the same token, the lifter who constantly overrates their RPE is likely to progress slower than necessary in the context of their training program. In the case of the latter, care should be taken to inquire about the trainee’s hesitation to add weight to the bar. For some, overrating RPE can be rooted in apprehension regarding prior injury or negative experiences in training history.
Despite the learning curve that RPE presents, evidence suggests that trainees can implement RPE to a reasonable degree of accuracy within the span of a few weeks. We can expect that an individual’s RPE rating may be a little wide of the mark at first. This is acceptable, as absolute accuracy is not required to reap the benefits of including autoregulation into the training approach. Over time, a trainee will hone the skill of accurately rating their RPE through practice and should be able to take into account the several factors that can determine the difficulty of a lift (e.g. technique, reps in reserve, velocity) and “wrap” these into an RPE rating. Barbell Medicine Coach Derek Miles provides a useful illustration in Figure 3.
What are the Practical Applications of RPE?
Let us now consider some examples of how one might use RPE in resistance training, beginning with the warm-up. In some athletic development contexts, warm-ups can be used for drill work, plyometrics, skill/coordination practice, and injury risk reduction (e.g., the FIFA 11+ sequence).
For a routine resistance training session, we recommend warm-ups that are specific to the planned training in nearly all cases. For a barbell training workout, this typically involves performing multiple sets of relatively low reps (e.g., 2-6 sets of 3-5 reps) of your first planned movement using the empty barbell, until you feel ready to increase load. Rest periods between these sets should be brief (20–60 seconds). Barbell “complex”-type warm-ups, wherein 3–5 complementary lifts are combined and performed with an empty barbell in a circuit fashion with no rest between movements (e.g., deadlift + overhead press + front squat), can also be used.
Weight Selection
Autoregulation with Multi-Rep Sets:
Consider a trainee, say, someone who has some background in strength training starting our Beginner Prescription program. The program calls for the trainee to work up to a set of 4 repetitions @ RPE 8 in the back squat:
Exercise protocol: Back Squat 4 reps @ RPE 6, 4 reps @ RPE 7, 4 reps @ RPE 8 x 1 set.
The lifter will warm up with the empty barbell in sets of 4 reps, taking 3 to 5 even jumps between the empty bar and their working set, as follows:
- Warm Up 1: 95 lbs x 4 reps x 1 set (minimal rest — typically, the amount of time it takes to load the bar for the next set)
- Warm Up 2: 135 lbs x 4 reps x 1 set (minimal rest)
- Warm Up 3: 185 lbs x 4 reps x 1 set (minimal rest)
- Final Warm-Up: 225 lbs x 4 reps x 1 set (rest 1 min)
- Working Set 1, based on “expected” effort required: 245 lbs x 4 reps x 1 set (rest 3–5 min)
The lifter rates this last set @ RPE 6, i.e., this is the first set to require noticeable effort, with an estimated 4 reps in reserve before failure. The lifter adds 5% (roughly 10 lbs) to the bar, as an addition of 5% in weight generally results in an increase in ~1 RPE at the same rep range (this varies across individuals and exercises):
- Working Set 2: 255 lbs x 4 reps x 1 set @ RPE 7 (rest 3–5 min)
- Working Set 3: 265 lbs x 4 reps x 1 set @ RPE 8
Let us assume that this same lifter has a training session the following week with the same prescribed protocol for the top set (i.e., 4 reps @ RPE 8). The lifter plans to increase the target weight anywhere from 1–5% based on this week’s performance. Let’s explore a few iterations of this:
Scenario A:
The lifter repeats the same warm-up scheme and finds that the final warm-up at 225 lbs feels about the same as the previous week. The lifter feels confident in progressing to 250 lbs, 260 lbs, then 270 lbs for their first, second, and final working sets, respectively. The lifter is stronger now than the previous week, having lifted a heavier load at the same relative effort.
Scenario B:
The lifter repeats the warm-up scheme but finds that the final warm-up at 225 lbs (and perhaps the one before this, too) feels particularly heavy on this day. In fact, the lifter estimates that they had about 3 reps left in the tank at 225 lbs (i.e., RPE 7). They decide that 235–240 lbs would bring them closest to their target of 4 reps @ RPE 8 on this day, so they adjust the load accordingly and complete the top sets before moving on to their next exercise.
Scenario C:
The lifter repeats the warm-up scheme and finds that the lifts are moving exceptionally well. The first warm-up at 95 lbs feels like an empty bar, and the last warm-up at 225 lbs barely registers on the RPE scale. On this day, the lifter goes for a substantial increase. They jump to 255 lbs, 265 lbs, then 275 lbs for their first, second, and final working sets, respectively — a 10-lb, or 4% jump from their previous exposure.
Note that in every scenario the warm-ups and work-up sets help to dictate the weight selection. The lifter does not execute a planned progression and then rate the RPE after the fact. Instead, they use real-time feedback to prospectively adjust the weight on the bar during any given training session.
Autoregulation with Singles:
We now turn to another lifter. This lifter has a long and consistent training history and they have just come off of a cycle wherein they performed high-RPE top sets of 5 reps in the deadlift. The protocol in their new phase of training is as follows:
Exercise protocol: Deadlift 1 @ 8, 75% e1RM x 4 reps x 6 sets
Previous 1RM: 425 lbs
Instead of using performance data from their last training phase to extrapolate weight selection for this new phase of training, the lifter slowly works up to find their prescribed single @ RPE 8. Here, the lifter uses a typical pyramid rep scheme to warm up:
- Warm-Up 1: 135 lbs x 5 reps x 1 set (minimal rest)
- Warm-Up 2: 225 lbs x 3 reps x 1 set (minimal rest)
- Warm-Up 3: 315 lbs x 2 reps x 1 set (rest 1–2 min)
- Warm-Up 4: 355 lbs x 1 rep x 1 set (rest 3 min)
- Final Warm-Up: 375 lbs x 1 rep x 1 set (rest 3 min)
The final warm-up moves fairly quickly, like an RPE 7 or so. The lifter decides to jump 5% for their single @ 8:
- Top Set: 395 lbs x 1 rep x 1 set.
The lifter uses the RPE percentage chart to calculate an estimated 1RM of 428 lbs. The lifter will use this to benchmark their progress throughout this phase of training and to calculate the load for their proceeding back-off sets. They will plan to improve on the single in the next exposure, while accounting for RPE.
With singles, especially, it can be tempting to ignore feedback on your last warm-up (the “single before the single”) and “bloat” the estimated one-rep max. Having the RPE climb on the singles from week to week could have negative downstream effects on performance and/or fatigue management, assuming the program is not designed to sharply escalate the RPE (for example, in a taper or peaking situation). In the best-case scenario, the lifter will “peak” prematurely; in another instance, fatigue quickly outpaces fitness development and performance consequently decreases.
Autoregulation with New Exercises
Consider a lifter performing an exercise they have never performed before, for example here, the front squat. It is in the lifter’s best interest not to extrapolate weight selection using arbitrary ratios based on performance in a related exercise (in this case, the back squat or the clean, for example). Rather, the lifter should work up slowly to the prescribed rep and RPE pairing for the programmed exercise. When performing an exercise for the first time, the lifter can increase the number of warm-up sets at the prescribed rep range and/or take smaller jumps in weight until reaching the intended RPE. The protocol here is as follows:
Exercise Protocol: Front Squat 6 reps @ RPE 6, 6 reps @ RPE 7, 6 reps @ RPE 8 x 1 set.
Given the lifter’s training history, which includes a 250-lb back squat (a lift with similar movement pattern to the front squat), they can expect some carry-over in strength to this lift. The lifter’s workout may proceed as follows:
- Set 1: 40 kg x 6 reps (minimal rest)
- Set 2: 50 kg x 6 reps (minimal rest)
- Set 3: 60 kg x 6 reps (minimal rest)
The lifter’s third set at 60 kg feels akin to a warm-up, so they take another 10-kg increase to in the fourth set, which they rate @ RPE 7:
- Set 4: 70 kg x 6 reps (rest 3 min)
They make one additional increase for their top set and stop here:
- Top Set: 75 kg x 6 reps
The lifter does not go back and make up the set @ RPE 6, although they may take smaller jumps to work up in their next training session with this protocol.
We hope that this guide was a helpful introduction to autoregulation and makes for better understanding of the whys and hows of using RPE in your training. For more information regarding the use of RPE in resistance training, check out Barbell Medicine Coach Alan Thrall’s two-part video series, linked here and here. Happy (autoregulated) training!
Thanks to Austin Baraki, MD and Tom Campitelli for his assistance in editing this article.
