Bench Press RPE 8 for 5 Reps: The Percentage Walkthrough

A bench press programmed as "5 reps at RPE 8" is the most common rep-and-effort prescription in modern strength training. Translating that into a number on the bar requires either the lifter's intuitive calibration or a table. The table is more reliable, especially early in a training cycle. Here is what the converter returns and how to read it.

The scenario

A trained intermediate bencher with a tested one-rep max of 140 kg. Programming calls for 3 working sets of 5 reps at RPE 8 on the day. Question: what weight goes on the bar?

What the calculator returns

Running the inputs through the RPE to Percentage Converter:

# rpe-to-percentage-converter (computed live from /engines/rpe-to-percentage-converter.js)
Engine input
  rpe                   = 8
  reps                  = 5
  one_rm                = 140

Engine output
  selectedPercentage    = 81.1
  estimatedWeight       = 113.5
  table (108 items)     = [...]

Selected percentage 81.1, working weight 113.5 kg. The converter prints the full 12 × 9 RPE-to-percentage table for reference. The RPE 8, 5-rep row at 81.1% is one cell in a larger lookup; the table allows quick re-reading for any rep/RPE combination across a typical training week.

Reading the numbers

The RTS lookup is symmetric across reps and RPE. Adjacent cells in the table illustrate the gradient:

5 reps × RPE 10        86.3%   (true 5RM — no reps in reserve)
5 reps × RPE 9         83.7%   (one rep in reserve)
5 reps × RPE 8         81.1%   (two reps in reserve)    ← target
5 reps × RPE 7         78.6%   (three reps in reserve)

For 140 kg 1RM:
  RPE 10 / 5 reps   = 120.8 kg
  RPE 9  / 5 reps   = 117.2 kg
  RPE 8  / 5 reps   = 113.5 kg  ← prescribed weight
  RPE 7  / 5 reps   = 110.0 kg

The 2.5 to 3.5 kg gap between adjacent RPE values is the practical precision the lifter has to work with. For a 5-rep set, the difference between RPE 7 (warm-up volume) and RPE 8 (working volume) is roughly one rep in reserve — and the bar weight gap is about one standard "small plate" jump.

Across a typical four-week block running this prescription, the lifter would build from 3 sets of 5 at RPE 7 in week 1 (110 kg) to 3 sets of 5 at RPE 8 in weeks 2 to 3 (113.5 kg) and a top set at RPE 9 in week 4 (117.2 kg). The table covers all of these one-row lookups without re-deriving percentages by hand. The full 12 × 9 table also handles rep schemes like 3 at RPE 9 (90.7%, or 127 kg here) and 8 at RPE 8 (73.9%, or 103.5 kg) — useful for back-off sets after the working singles or triples.

Where the table breaks

Three failure modes show up.

1RM is stale. The percentage anchors everything; an out-of-date 1RM input invalidates every cell. If the lifter's true bench has moved to 145 kg in the months since the test, RPE 8 / 5 reps is closer to 117 kg than 113.5 kg. Re-test or use the One Rep Max Calculator on a recent clean 3RM every 8 to 12 weeks.

RPE calibration drift. New lifters consistently rate sets one point lower than trained populations (Hackett 2017)^[2]. A lifter who calls a set "RPE 8" but actually had four reps in the tank is training closer to 78% — the table loses validity. Cross-check with bar speed: at RPE 8 for 5 reps, the bar should noticeably slow on rep 4 and rep 5.

Lift-specific differences. Deadlift RPE calibrates differently than bench. The RTS table was originally fit on Tuscherer's powerlifting cohort and assumes similar bar-path dynamics across the squat, bench, and deadlift. Olympic-lift derivatives and tempo work do not map cleanly to this table; use velocity-based methods instead.

Velocity cross-check

For lifters with a barbell velocity tracker, 81.1% of 1RM in the bench press should produce roughly 0.40 to 0.55 m/s on the concentric phase of the first rep. The Velocity Based 1RM tool maps the inverse: from observed bar speed and weight, estimate the actual percentage of 1RM. Used alongside the RPE table, the velocity check catches stale 1RM inputs in real time.

A practical session example: lifter sets up 113.5 kg, presses 5 reps with rep 5 noticeably slow but smooth (no missed lockout, no chest grind). First-rep velocity reads 0.48 m/s. That confirms RPE 8 calibration. If rep 1 had read 0.65 m/s, the lifter was actually closer to RPE 6 to 7, and either the 1RM input is too high or the lifter is having an exceptional day. If rep 1 read 0.32 m/s and the set crawled, the lifter is at RPE 9 to 10 and should drop the planned remaining sets to 5 percent below the prescribed weight.

The velocity-RPE-percentage triangle is the strongest auto-regulation framework in current strength programming. Any two of the three readings catch errors in the third.

Related tools and follow-ups

• RPE to Percentage Converter — the converter used in this walkthrough.
• One Rep Max Calculator — recalibrate the anchor every 8 to 12 weeks.
• Velocity Based 1RM — independent percentage cross-check from bar speed.

For broader context: How to use RPE for training, RPE-based programming: the math vs the coach, and Velocity-based training: 1RM from bar speed cover the broader programming options.

FAQ

What weight is RPE 8 for 5 reps on a 140 kg bench press? 113.5 kg. The RTS table maps RPE 8 at 5 reps to 81.1 percent of 1RM. 0.811 times 140 equals 113.54, which rounds to 113.5 kg on a standard 0.5 kg precision plate set.

What does RPE 8 actually mean? RPE 8 means the set ends with about 2 reps left in reserve. On the 10-point Reactive Training Systems scale, RPE 10 is a maximal grinder, RPE 9 has one rep left, RPE 8 has two, and RPE 7 has three.

How reliable is RPE as a load-selection method? Hackett 2017 found intra-class correlation coefficients of 0.83 to 0.89 for RPE in resistance training. That is reliable for trained lifters; novices and lifters new to the scale tend to under-rate effort and accumulate fatigue across sessions.