How To Do the Lateral Raise

Start here

The lateral raise is the most-butchered exercise in any gym, and the reason is simple: it is the one where ego costs you the most and buys you the least.

It is a small muscle moving a long lever. That combination means the weight you can move with good position is embarrassing, and the weight you can move with body english is not.

Everything below is about resisting that, plus a study that took the cable-versus-dumbbell argument seriously enough to run it on the same person's two arms.

The study that used each lifter as their own control

This is the cleanest design you could ask for on this question.

Twenty-four resistance-trained men and women completed an eight-week intervention of two weekly sessions, each five sets of lateral raises taken to momentary failure. It used a within-participant design: each participant's arms were randomly allocated, one to the cable lateral raise and one to the dumbbell lateral raise.

Same person. Same nutrition, sleep, genetics, training history. The only difference between the two arms was the implement.

Critically, shoulder adduction and abduction range of motion was standardised and matched between the conditions, so the comparison was not contaminated by one version being performed through a longer arc than the other.

Muscle thickness of both the proximal and distal lateral deltoid was measured by B-mode ultrasound, and the data were analysed in a Bayesian framework with univariate and multivariate mixed effect models, with differences estimated as average treatment effects and inferences based on posterior distributions and Bayes Factors.

Lateral deltoid muscle thickness increased by 3.3 to 4.6 percent across the intervention.

That is your realistic expectation for eight weeks of hard, dedicated side delt work. Roughly three to five percent thickness, from ten sets a week taken to failure. Not nothing, and not the transformation the internet promises.

The paper reports moderate support in its univariate analysis for a difference between the conditions, quantified with Bayes Factors. I am not going to tell you which implement won, because the direction and magnitude of that comparison sit in the full paper rather than in what I can verify here, and guessing it would be worse than useless.

Twenty-four participants over eight weeks. What it does establish confidently is the growth range from that dose, and that the question was tested properly rather than argued.

Why the implement changes anything at all

The mechanism behind the cable-versus-dumbbell argument is the resistance profile, and that is worth understanding whichever you pick.

A dumbbell is hardest where your arm is horizontal, because that is where gravity has the longest lever on it, and almost weightless at the bottom where your arm hangs by your side. A cable pulling from low and across keeps tension on at the bottom, where a dumbbell gives you nothing.

A review examined the interplay between muscle length, range of motion and exercise selection, synthesising long-term studies on isometric training at different muscle lengths, on training with different ranges of motion, and on exercises that load muscles at different lengths and with different resistance profiles.

Different lines of investigation suggest that muscle length plays a role, with training at longer muscle lengths eliciting more favourable muscle growth.

For the lateral raise, the longer muscle length is the bottom, arm down by your side. That is exactly the position where a dumbbell provides the least resistance and a cable provides the most.

That is the whole mechanical argument for cables in one sentence.

The review is careful about its own limits and so am I: it states that it is difficult to identify whether muscle length itself is the main factor, which limits the practical conclusions available. This is the direction the evidence points, not a proven rule, and the trial above standardised range of motion between its conditions precisely because these variables are hard to separate.

The setup

Stand tall, feet about hip width, a small bend in the knees.

A slight forward lean from the hips is fine and many lifters prefer it, because it puts the arc a little more in front of the body. Pick one and keep it, rather than leaning further as the set gets hard.

Arms slightly bent at the elbow and kept at that same angle throughout. A bend that opens and closes during the set is a way of cheating the lever shorter, and it means every rep is a different exercise.

The most important cue: brace your midsection and stop your torso moving. If your body is swinging to launch the weight, the deltoid has been demoted to a passenger.

Standard coaching practice rather than a finding from the sources cited here.

The movement itself

Lead with the elbow, not the hand. If your hand rises first, you are rotating the movement into something else and letting the arm turn into a lever you cannot control.

Raise to roughly shoulder height. Higher than that hands the work increasingly to the traps, which is a legitimate choice if that is what you want, but it is a different goal.

Do not shrug. Keep the shoulder itself down as the arm goes up. The trap wants this job and it will take it if you let it.

Think about pushing your hand away from your body rather than lifting it up. The arc widens, the lever lengthens, and the delt does more with less weight.

Control the way down. This is where most people give the set away entirely: a hard raise followed by a free-fall. Lower under tension across two seconds or more.

And on the load, be honest. If you cannot start the rep without a hip drive, the weight is wrong. The lateral raise is the one exercise where going lighter reliably makes it harder.

Also standard coaching practice, included because a guide that skipped it would be useless.

How much side delt work is realistic

For a sense of what serious volume looks like, there is data on people whose job is their physique.

A study quantified and compared weekly training volume per muscle group in 154 competitive physique athletes across different federations and categories, on the grounds that specific literature at this level is scarce, then compared their practice against published guidelines.

What the data shows is variation rather than a magic number: many categories reduced training volume in the pre-contest period for most muscle groups, and the average number of sessions per week per muscle group differed between off-season and pre-contest. Significant differences in weekly set volume appeared between categories, with deltoid volume among the specific differences identified, including reductions in some categories and differences between categories in the off-season.

The useful conclusion is not a number to copy. It is that even among people optimising full-time, delt volume is deliberately varied across the year rather than pushed permanently upward.

The trial earlier used ten hard sets a week, two sessions of five sets to failure, and produced 3.3 to 4.6 percent thickness change in eight weeks. That is a concrete, tested dose you can actually run.

This is self-reported survey data from competitive physique athletes describing their own training, not a controlled comparison of volumes, so it describes practice rather than establishing what works.

Why any of this is worth your time

Side delts are the muscle that changes your silhouette more than almost anything else per unit of effort, and they are trained badly almost universally because the correct version looks unimpressive.

What the evidence supports: eight weeks of ten weekly sets to failure moved lateral deltoid thickness by roughly 3 to 5 percent. The cable-versus-dumbbell question is a genuine one, tested within the same person, with the resistance profile at the bottom of the range being the mechanism people argue about. And training at longer muscle lengths appears favourable, which is the bottom position here.

What that leaves you: lighter than your ego wants, elbow leading, torso still, shoulder height, controlled on the way down, and enough sets across the week to matter.

This is a training article, not medical advice. Shoulder pain during raises is a reason to see a qualified professional, not a reason to adjust your grip and push on.