Incline dumbbell press: logging, progression and the ladder

An incline dumbbell press is a dumbbell press on a bench set between flat and upright, usually around 30 degrees, and in the app's library it trains the upper chest and front delts first, with the triceps second. The number you log is the weight of one dumbbell. A lifter who presses the 32s for ten at RPE 8 has done a set worth an estimated 47.1 kg for one per hand, computed with the model behind the one-rep max calculator. The awkward part is progression: the next pair up is 34 kg, a 6.3 per cent jump, so the reps have to do most of the climbing.

Every figure below is compute.mjs output, the script that runs the app's own formulas, or it carries a citation to the study it came from.

What the incline dumbbell press trains

The library entry is Dumbbell Incline Press: a compound horizontal push, primary muscles upper chest and front delts, secondary triceps. The app counts weekly sets on a 14-muscle grid with no separate upper chest, so the upper chest is counted as chest. Each working set credits the chest and front delts one set apiece and the triceps half a set.

The angle is why the lift exists. Thirty trained adults pressed at 0, 15, 30, 45 and 60 degrees at 60 per cent of their 1RM, and activity in the upper portion of the pectoralis major peaked at 30 degrees [1]. The middle and lower portions were most active on the flat bench. The anterior deltoid peaked at 60 degrees, and past 45 degrees the authors saw the chest's performance fall away [1]. A bench set two notches too high turns the lift into a shoulder press with a backrest.

Activity is not growth, and the growth evidence is thinner. In a study of 47 untrained men training once a week for eight weeks, the incline group grew most at the upper chest site, 0.62 cm more than the flat group there, while the other two thickness sites showed no difference between groups [2]. Untrained men, one session a week, eight weeks: treat it as a hint, not a rule.

How to log it: one dumbbell, every time

Type the weight stamped on one dumbbell. A pair of 32s is logged as 32 kg, not 64 kg. The app stores what you type and cannot know whether you meant one hand or two, so what matters is that every session uses the same convention. Per hand matches the rack and every other dumbbell lift in your log.

Both arms press together, so log one line per set, not per side. The per-side flow, left then right, is for unilateral lifts like the single-arm row.

Rate every working set by reps in reserve: RPE 10 is nothing left, 9 is one rep left, 8 is two [3]. A set saved without RPE is treated as RPE 10, so 32 kg for ten with the field blank estimates 43.3 kg instead of 47.1 kg. Epley and Brzycki both put the same set at 42.7 kg, because they assume the tenth rep was the last one available.

Warm-ups never count toward volume. Two sessions a week of three working sets, each with one warm-up in front, come out of the app's volume code as 6 chest sets, 6 front delt sets and 3 triceps sets. Each extra weekly set was linked to about 0.37 per cent more growth in a meta-analysis of 15 studies [4], and the post on weekly sets per muscle covers how far to push that.

The ramp and the ladder for 32 kg

The app's warm-up builder is written for a barbell. It starts from a 20 kg bar and adds percentage steps only once the working weight passes 50 kg, so for 32 kg it returns a single warm-up set. On dumbbells, read the bar as a pair of 20s.

Ten reps with two left means 32 kg was a 12-rep load that day. The model reads the reps-by-RPE chart along its diagonal, reps plus reps in reserve, and lands on 68 per cent of 1RM. The estimated 1RM guide explains how the chart is read. The ladder below, from compute.mjs ladder --kg 32 --reps 10 --rpe 8, gives the heaviest load per hand for each rep count at RPE 10 on that day.

Nobody should test a single on dumbbells, and the top rows are the least trustworthy part of the table. Its use is the bottom half. The next pair, 34 kg, sits between the 10-rep max of 34.8 kg and the 11-rep max of 33.3 kg. Ten clean reps with the 34s next week would be close to a true max, which is the whole progression problem in one line.

Why the 2 kg step needs a rep range

The app's step for a dumbbell lift is 2 kg, because that is how most racks go up. On a 32 kg dumbbell that is a 6.3 per cent jump, according to the app's own next-weight rule. A load jump that size cannot be repeated at the same reps every week. So the reps carry the progression and the load moves only occasionally.

Of the seven progression schemes (Auto, Double, Top set + back-off, Load, Reps, RIR and Manual), Double suits this lift. Set a range of 8 to 12. Hold 32 kg and add a rep where you can, at RPE 8. When all three sets reach twelve, the rule adds the 2 kg and resets the target to eight. The post on double progression has the full logic.

Here is what that looks like in numbers. Three sets of 32 kg for twelve at RPE 8 are worth 49.3 kg each. The first week on the 34s, eight reps at RPE 8 is worth 46 kg. The estimate falls by 3.3 kg on the day the lift got heavier, because twelve reps with two left rates higher than eight reps with two left on this curve. Compare like with like instead: 32 kg for ten at RPE 8 was worth 47.1 kg, and 34 kg for ten at RPE 8 is worth 50 kg. The jump is real once the reps come back.

The app's next-weight rule reads the rating as well as the reps. With a target of ten reps at RPE 8, a set of 32 kg for ten at RPE 8 holds at 32 kg. At RPE 7 it still holds, because one rep of slack does not earn a 6.3 per cent jump. At RPE 6 it suggests 34 kg. At RPE 9.5 it drops to 30 kg. After 35 days away from the lift, the same 32 kg for ten comes back as a 30 kg suggestion, the detraining rule for any gap over 30 days.

A deload week never emits an increase, and a load increase never leaks through on a cut. Three sessions without progress flag a stall by default.

Which mistakes break the numbers?

The first is switching conventions. One session logged as 32 kg and the next as 64 kg turns the history into noise, and the estimate doubles overnight. Decide on per hand and keep it.

A moving bench does the same damage. Benches have notches, not degrees, and activity shifts with the angle [1], so a higher notch is a different lift with a different rep count. Note the notch, or log a deliberately different angle as a new exercise.

Grinding to the top of the range is the quiet one. Twelve reps at RPE 10 on 32 kg is worth 47.1 kg, exactly what ten at RPE 8 was worth. That twelfth rep earns the jump without the strength behind it. Count it only when it comes with two in reserve.

What changes when you swap it

The barbell incline bench press is in the library with chest and front delts as primary muscles and triceps as secondary, so the volume credit is the same. The step becomes 2.5 kg on a much bigger number, which is a smaller jump in percentage terms. Expect more load: in 12 resistance-trained men, flat-bench chest press 1RMs were 17 per cent lower with dumbbells than with a barbell [5]. Chest and front delt activity did not differ between the two, while triceps activity was lower with dumbbells. Never carry a dumbbell number across. Start the barbell history fresh, and see the bench press progression post for how a barbell press moves.

The machine incline press lists only the upper chest as primary and the triceps as secondary, so it credits no front delt sets at all. Its step in the app is 1 kg. The Smith machine incline bench press lists upper chest and chest as primary and front delts and triceps as secondary, so each set gives the front delts half a set instead of a full one.

Expect every incline variation to be weaker than its flat version. In 12 national and international bench press athletes, 6RM loads were 18.5 to 21.5 per cent lower on the inclined bench than on the flat or declined bench [6]. That gap is the angle doing its job.

Questions

Do I log both dumbbells or one for the incline dumbbell press?

Log one. A pair of 32s is 32 kg in the log. The app records the number you type and cannot tell a pair from a single, so consistency is what matters, and per hand matches the rack and every other dumbbell lift. Entering 64 kg one week and 32 kg the next makes the estimated 1RM jump and fall for no reason.

What bench angle is best for the incline dumbbell press?

Around 30 degrees. In a study of 30 trained adults, upper chest activity peaked at 30 degrees, and past 45 degrees the front delts took over and chest performance fell. Gym benches set by notches, so find the notch closest to 30 degrees, note it, and use the same one every session so the reps stay comparable from week to week.

How do I progress an incline dumbbell press in 2 kg steps?

Use a rep range. Hold the weight across a range such as 8 to 12 reps at RPE 8, and move to the next pair only when every working set reaches the top. A 2 kg step on a 32 kg dumbbell is a 6.3 per cent jump, too large to take every week, so expect to start the heavier pair near the bottom of the range.