What RPE should a top set be for hypertrophy and strength?

A top set should be RPE 8 most weeks: two reps left when you rack it. Push it to RPE 9 in the last week or two of a block, and take it to RPE 10 only when testing a max. That holds for hypertrophy and strength alike, since the reps change between the two goals and the effort does not. The log shows why the rating matters. A 90 kg set of six at RPE 8 is worth about 114.5 kg for one in the app's model, and the same six reps ground out at RPE 10 are worth 107.5 kg. The grinder is not a stronger set. It is a weaker day.

Why RPE 8 is the default

The scale the app uses puts RPE 10 at failure, 9 at one rep left and 8 at two left, down through RPE 6 at four left. The coaching review that laid out how to train with it recommends 6 to 12 reps at RPE 8 to 10 for hypertrophy, depending on the phase, with failure kept to the last set for a muscle and mostly to isolation lifts. It suggests main lifts such as squats and presses stay in RPE 6 to 8 for most of the training, so that fatigue does not eat the sets that follow [1].

For strength the same review recommends the 1 to 6 rep range, and its chart shows that 6 reps at RPE 10, 5 at RPE 9, 4 at RPE 8 and 3 at RPE 7 all land near 83 per cent of a max, roughly a 6RM [1].

RPE 8 is where both goals meet. It is hard enough to count as near-failure work and leaves enough in reserve that the back-off sets after it are still worth doing.

An eight-week trial that matched sets and reps between an RPE group and a percentage group had the RPE group work through weekly targets of RPE 6 to 8, then 7 to 9, then 8 to 10, adjusting the load by 2 per cent for every half point a set missed its target [2]. Their squat rose 17.05 kg against 13.91 kg in the percentage group, and their bench 10.70 kg against 9.64 kg, differences that were not statistically significant, and their average squat RPE across the block was 7.2 [2]. Effort in that range built as much strength as the percentage chart did.

How a logged RPE changes your estimated 1RM

The app has one estimated 1RM model: the reps-by-RPE chart read along its diagonal, where reps plus reps in reserve decide what percentage of a max the set represents. The one-rep max calculator runs the same code. For a 90 kg set of six, the RPE you log moves the estimate like this.

A set logged without an RPE is treated as RPE 10, the assumption every classic formula makes: Epley gives 108 kg for those six reps whatever you had left. Log the top set honestly and the estimate tracks your strength. Leave RPE blank and every top set looks like a max, so the trend flattens and the next weight never moves. The estimated 1RM guide covers the model in full, including what happens past the end of the chart.

RPE 9.5 reads as 0 to 1 reps left: the set you were not sure you could repeat. It is a real rating, and a top set that lands there in the last week of a block is doing its job. The same rating in week one means the load is too high for the plan.

Inside a block the app plans top sets on an RIR ramp from 3 down to 1, which is RPE 7 rising to RPE 9. Week one at RPE 7 feels light, and it is meant to. By the last week the same rep target at RPE 9 should sit under a heavier bar, and the rep-max ladder from a 90 kg six at RPE 8 says roughly how much: 98.8 kg for five, 105.6 kg for three, 114.5 kg for one.

When should a top set be RPE 10?

When you are testing, and rarely otherwise. In the study that validated the scale, 15 experienced squatters rated their true one-rep max 9.80 on average while 14 novices rated the same lift 8.96, and the experienced group moved the bar at 0.24 metres per second against 0.34 for the novices [3]. A novice who calls a set RPE 8 may have four reps left, not two. The same study found that an eight-rep set at 70 per cent of a max drew ratings from 6.5 to 10 across the experienced lifters [3]. A percentage cannot set the effort of a top set. Only the rating can, and it needs practice.

The practical test from the coaching review is this: take a set a few reps short of failure, log the RPE, then repeat the load to failure and see whether the reps you had left matched the number you gave [1]. Until the two agree, log RPE on every set and keep the top set planned at RPE 8, where a rating that is one point off costs you a rep rather than a week.

Questions

Is RPE 8 the same as two reps in reserve?

Yes. On the scale the app uses, RPE is 10 minus reps in reserve, so RPE 8 is two reps left, RPE 9 is one and RPE 10 is failure. Half points fill the gaps: RPE 8.5 is one to two left and RPE 9.5 is zero to one. The estimated 1RM model uses the same arithmetic, which is why the RPE you log changes the number it shows.

Should the top set be a higher RPE for strength than for hypertrophy?

No. Both sit at RPE 8 most weeks, and the reps change rather than the effort. A strength top set is a triple or a five at RPE 8, and a hypertrophy top set is eight to twelve reps at the same rating. Both move to RPE 9 late in a block. The one difference is that a strength block ends with a test, and a test is the one planned RPE 10.

What happens if I log a top set without RPE?

The model treats it as RPE 10, a true max. For a 90 kg set of six that is 107.5 kg instead of the 114.5 kg an honest RPE 8 would give, so you look 7 kg weaker than you were. Do it every week and your estimated 1RM trend flattens while you are getting stronger. Log the RPE, even a rough one; a rough rating is closer to the truth than a blank.

When should a top set go to failure?

Rarely on a main lift. The coaching research limits failure to the final set for a muscle, mostly on isolation lifts, because a set to failure can cut what you do on the sets after it. On squats, presses and pulls, RPE 9 to 9.5 in the last week of a block is as far as a planned top set should go. RPE 10 is for the day you test.