Bicep Curl Variations: Types of Curls and When to Use Them

How curl variations actually differ

All bicep curl variations bend the elbow against resistance. What changes is where the curl is hardest, where your upper arm sits and which way your palm faces. Those three factors, plus how stable and loadable the lift is, tell you more than the name.

Resistance profile comes first. A dumbbell is hardest when your forearm is horizontal, so the angle of your upper arm moves the hardest point; a cable is hardest where it pulls at right angles to your forearm. An EMG study in 22 trained men found preacher curls worked the biceps long head harder than standing and incline curls in the first third of the lift and less in the final third.

Shoulder position comes second. The biceps crosses the shoulder, so curling with the arm in front of you (preacher, spider, concentration) shortens it, and curling with the arm behind you (incline, Bayesian) lengthens it. The long-head guide and short-head guide cover what that means for each head.

Grip comes third. The biceps also turns the palm up, so palms-up curls suit it best, while hammer and reverse grips likely shift more of the work toward the brachialis and brachioradialis, as the brachialis exercises guide explains.

Bicep curl variations compared

These are the main types of bicep curls. The hardest point assumes your upper arm stays where the variation puts it; if your elbows drift forward, the hardest point moves earlier.

Pick two rows that differ in hardest point or shoulder position. A preacher curl and an incline curl load the arm in different ways, while a barbell curl and an EZ-bar curl are nearly the same lift. At home, the dumbbell biceps guide builds a plan from dumbbells alone.

Free weights vs cables: the resistance-profile question

Cables are often sold on constant tension, but the real difference is where each tool loads the muscle. A 10-week trial in 35 untrained young adults compared cable and barbell preacher curls, which load the biceps hardest at opposite ends of the range. Biceps thickness grew about the same: 7% with cables and 8% with the barbell.

The barbell group gained more strength near the straight-arm position (39% versus 30% at 20° of elbow flexion), while strength at 60° and 100° improved similarly. So choose by comfort, setup and how easily you can add small amounts of load: cables make small jumps and odd angles easy, while dumbbells and bars need no station.

One arm or two seems to matter little for growth. A 2025 meta-analysis of nine studies, not specific to the biceps, found one-limb and two-limb training built similar muscle, with strength gains matching the way people trained.

An EMG study in 12 trained men found EZ-bar and straight-bar curls produced somewhat higher biceps activity than dumbbell curls, with little difference between the two bars. EMG measures electrical activity during the lift, not growth, and a 2022 review concluded it is not a validated predictor of muscle growth.

Partial reps and the stretched position

The bottom half of the curl, where the biceps is stretched, has the better growth data. An 8-week trial in 19 untrained women trained one arm only in the bottom half of a dumbbell preacher curl (0–68° of elbow flexion) and the other only in the top half. The bottom-half arm grew more at the lower biceps site, though not at mid-arm, and gained more full-range strength.

A 5-week trial in 32 untrained adults found the same pattern: stretched-range preacher partials (0–50°) thickened the elbow flexors 8.9% versus 3.4% for shortened-range partials (80–130°). A systematic review of 11 studies concluded that full-range curls and stretched-range partials tended to produce more lower-arm growth in the biceps and brachialis than top-range partials.

Those trials did not test stretched partials against full reps. In trained lifters, an 8-week within-participant trial found lengthened partials and full-range reps grew the upper arm similarly. A pre-registered trial across 15 sites, with 163 lengthened-partial and 134 full-range participants, found differences in estimated arm size that were small and likely practically equivalent over 12 weeks.

Full range still wins for strength. A 10-week trial in young men found full and partial elbow-flexion training both grew the muscle (9.65% and 7.83%), but full range raised the 1RM more (25.7% versus 16.0%). Use full reps by default, and if you add partials, keep them in the bottom half.

21s, drag curls and other named curls: what is known

A set of 21s is seven reps in the bottom half, seven in the top half and seven full reps, without rest. No trial has tested them. Given the partial-range trials above, the top-half block is probably the least productive part, so treat 21s as an occasional finisher rather than a better way to curl.

Drag curls keep the bar against your torso while the elbows travel back, so the shoulder extends as you lift. No trial has tested them either, and you will use less load than in a strict curl. They are a variety option, nothing more.

The Bayesian curl is a cable curl with the arm behind the body, often promoted for the long head. A 10-week trial in 15 untrained men matched the cable resistance profile and found biceps growth of 7% with cable preacher curls and 9% with Bayesian curls, a difference that was not significant. A preregistered 10-week trial in 24 untrained men found extra shoulder extension did not add elbow-flexor growth, even though participants lifted more in that position.

Drop sets mainly save time. A small 8-week pilot in 9 untrained men found curls to failure grew the elbow flexors similarly with heavy sets at 80% 1RM, light sets at 30% 1RM or drop sets, which took the least time.

Which curl variation is best?

No single curl is best; for the wider ranking, read what the research says about the best biceps exercises. In a 12-week MRI trial in 21 untrained adults, total elbow-flexor growth was similar with incline and preacher curls (17.6% versus 19.2%), though incline curls grew the biceps slightly more and preacher curls the brachialis and brachioradialis.

An 8-week trial in 63 young women found incline curls grew the upper elbow flexors slightly more (0.08 cm) and preacher curls the lower part slightly more (0.10 cm). Small regional differences like these argue for pairing two curls, not crowning one.

Ranks reflect growth data first, then how easily you can progress the lift. Any two of the top three make a strong pair.

Technique cues for the core curls

• Preacher curl: set the pad high under your armpit, lower to an almost straight elbow and never bounce out of the bottom.
• Incline dumbbell curl: set the bench around 45–60°, let your arms hang straight down and keep the elbows still.
• Standing cable curl: keep your elbows at your sides, stand tall and finish without leaning back.
• EZ-bar curl: brace your trunk, grip about shoulder width and end the set when you need hip drive.

How many variations per workout

One or two curl variations per session are enough, and two or three across the week cover the main resistance profiles. A 2022 systematic review of 8 studies in young men suggests planned variation may help develop different regions of a muscle, while frequent random changes may hinder progress.

Choose one curl that is hardest in the stretched position, such as a preacher or incline curl, and one that loads the middle or top, such as a cable or standing curl. Keep the same pair for 6–8 weeks or longer so you can see whether your numbers rise.

Sets, reps and weekly volume

A 2026 meta-regression of 67 studies found growth kept rising with weekly sets, with diminishing returns, and little evidence that spreading the same sets over more sessions added growth. It counted indirect sets, such as rows for the biceps, as half a set because that method fit the data best. The sets per muscle guide explains how to count.

• Start with 6–12 direct curl sets per week over one or two sessions.
• Count each set of rows, pulldowns or chin-ups as about half a set.
• Use 6–12 reps for bar and preacher curls and 10–20 for cables and machines.
• Add sets only when progress stalls and you recover well.

Train hard enough to matter. Many of the trials above took sets to or near failure; the shoulder-extension trial, for example, used sets of 10–20 reps to momentary failure. Ending most sets 0–2 reps short of failure is a sensible default, and the RPE and RIR guide and RPE chart show how to rate that.

A sample arm day using two variations

Run this once a week, and on another day add 3–4 sets of standing cable curls or hammer curls after your back work, for 9–10 direct curl sets in total. The arm workout template shows the full session, and the biceps hub lists alternatives if a station is taken.

Make each pick easier or harder

Use double progression: reach the top of the rep range on every set, then add the smallest load step. The progressive overload calculator maps the steps.

Common mistakes

• Swinging your hips or leaning back, which shifts the hardest point and takes load off the biceps.
• Cutting off the bottom of the range, the part with the strongest growth data.
• Dropping the weight on the way down. In a 5-week trial in untrained adults, eccentric-only curls grew the elbow flexors as much as normal curls (9.7% versus 10.6%), while concentric-only curls produced no significant increase.
• Rotating to a new curl every session, which makes progress impossible to track.
• Picking curls by EMG rankings or catchy names.
• Pushing through sharp elbow or shoulder pain; stop and see a qualified clinician.

What curl research has not tested yet

• No trial has tested 21s or drag curls.
• Most curl trials use untrained people for 5–10 weeks.
• The two matched-profile cable trials do not support a Bayesian-curl advantage.
• No biceps-specific trial has compared one-arm with two-arm curls; only a general meta-analysis exists.
• Few studies separate the two biceps heads, so head-specific growth claims remain largely untested.

Until better data arrive, keep the stretched range in every set and pick two curls that differ. Let your logged numbers decide the rest.