Dumbbell Shoulder Press

What it trains. The dumbbell shoulder press is the overhead press with a weight in each hand. You press from shoulder height to straight arms overhead, seated or standing.

• Primary: the anterior deltoid raises the upper arm forward and up (shoulder flexion).
• Secondary: the medial deltoid helps lift the arm; the triceps straighten the elbow; the upper trapezius and serratus anterior rotate the shoulder blade upward so the arm can reach overhead; the clavicular (upper) head of the pectoralis major helps early in the press.
• Stabilisers: the rotator cuff keeps each upper arm centred in its socket while the dumbbell tries to wobble. Standing, the trunk also works to keep your spine still.

Dumbbells vs a barbell. Saeterbakken and Fimland (2013) tested 15 healthy men at 80% of their 1-repetition maximum (1RM) in four presses: seated and standing, with dumbbells and with a barbell.

• Anterior deltoid activity was about 11% higher with seated dumbbells than a seated barbell, and about 15% higher with standing dumbbells than a standing barbell.
• Triceps activity was about 39% higher with a standing barbell than standing dumbbells.
• Strength went the other way: the standing dumbbell press had the highest deltoid activity but the lowest 1RM: about 7% lower than the standing barbell press.

Kohler et al. (2010) found the same trade-off in 30 resistance-trained lifters pressing seated at their 10-repetition maximum. The less stable the setup, the less weight they could lift. Triceps activity was highest with a barbell on a bench and lowest with dumbbells on a Swiss ball. The abstract does not report deltoid differences between the setups.

So dumbbells are not a weaker shoulder exercise. You lift a little less and use a little less triceps, while each shoulder works to steady its own weight.

Seated vs standing. In the same 2013 study, with dumbbells:

• Medial deltoid activity was about 15% lower seated than standing, and posterior deltoid activity about 24% lower.
• Anterior deltoid activity was about 8% lower seated, but that difference was not statistically clear.
• The standing dumbbell press had a 1RM about 10% lower than the seated one.

For the trunk, Saeterbakken and Fimland (2012) measured the same kind of dumbbell press in 15 men. Seated, rectus abdominis activity was about 81% lower and external oblique activity about 58% lower than standing. Lower-back (erector spinae) activity was similar either way.

The trade is simple. Seated with a back pad, you can press a little more and your trunk has less to do, so it suits shoulder-focused work and lifters who find standing presses hard on the back. Standing asks more of the side and rear deltoids and the abdominal wall, for a little less load. Sitting on a Swiss ball instead of a bench added nothing: Uribe et al. (2010) found the same deltoid, chest and abdominal activity on both in 16 men.

Grip. Palms facing forward, elbows out to the sides and a little in front of your body, is the common version. Palms facing each other (a neutral grip) or turned halfway is a valid variant. It usually brings your elbows further in front of you, which some lifters find more comfortable. We found no study comparing grips in the dumbbell press, so neither is better or safer. Use the one that lets you press pain-free through a full range.

Range of motion. The ACSM 2026 position stand found that training through a complete range of motion improved strength more than partial range. Lower until the dumbbells are about level with your shoulders, or as far as stays comfortable, and press to straight arms.

Technique cues:

• Set the bench. Back pad upright, or one notch back from upright. Reclining it further turns the press into an incline press.
• Get the weights up safely. Sit with the dumbbells on your thighs, then kick each one up with your knee as you bring it to your shoulder.
• Set your base. Feet flat and wide, glutes and upper back against the pad. Standing: feet hip-width, glutes tight, trunk braced.
• Start position. Dumbbells at shoulder height, wrists stacked over your elbows, forearms roughly vertical.
• Press straight up. The dumbbells finish over your shoulders or just in front of your head, arms by your ears. They can come close at the top; they don't need to touch.
• Ribs down. Don't arch your lower back off the pad to finish a rep. That turns it into an incline press.
• Lower under control to the start. Breathe and brace before each rep (see Breathing Technique & the Valsalva Effect and Abdominal Bracing (vs Hollowing)).

• Beginners learning to press overhead, and anyone building shoulder strength and size.
• Lifters training alone. A failed dumbbell press is easy to bail out of: lower the weights to your shoulders or drop them to your sides.
• People whose shoulders feel better when each arm can find its own path, and lifters who want to even out a stronger and a weaker side.
• Lifters who find standing barbell presses hard on the lower back (use the seated version).

• Healthy shoulders and backs only. If raising your arm overhead is painful, you have a history of dislocation or recent shoulder surgery, or your lower back arches to get your arms up, get individual advice first.
• EMG is not growth. These studies measure short-term muscle activity in young, healthy, mostly male lifters; none compared muscle growth between press variations.
• Small studies. 15 to 30 participants each. They show direction, not exact numbers for you.

Next lesson: Lateral Raises. Train the side deltoid on its own and see where it fits beside the press.

• Primary: anterior deltoid (shoulder flexion). Secondary: medial deltoid, triceps, upper trapezius, serratus anterior, clavicular pectoralis major.
• Dumbbells raised anterior deltoid activity 11–15% over a barbell; the standing dumbbell press had the lowest 1RM, about 7% below the standing barbell and about 10% below seated dumbbells (Saeterbakken 2013, n = 15).
• Seated vs standing dumbbells: medial deltoid about 15% lower and posterior deltoid about 24% lower seated (Saeterbakken 2013).
• Seated pressing cut rectus abdominis activity by about 81% and external oblique by about 58%; lower-back activity was similar (Saeterbakken 2012, n = 15).
• Less stable setups meant lighter loads and less triceps activity (Kohler 2010, n = 30); a Swiss ball made no difference to muscle activity (Uribe 2010, n = 16).
• No study compared palms-forward and neutral grips in this press.
• Full range of motion builds more strength than partial range (ACSM 2026).