Bayesian Cable Curls: Form, Setup and Whether They Build Bigger Biceps

A Bayesian cable curl is a single-arm cable curl done with your back to the weight stack and the working arm slightly behind your torso. Doing the Bayesian cable curl form correctly means the biceps is stretched and under load from the first inch of the rep, instead of resting at the bottom the way it does in a standing dumbbell curl.

They are all over social media, so it is fair to ask whether they work. Short answer: the logic is sound, the direct evidence does not exist yet, and the exercise is a smart way to train the stretched end of the range. This page covers the setup, the cues, and what research can and cannot tell us. It sits inside our complete arm and forearm training guide.

What is a Bayesian cable curl?

Stand facing away from a low cable pulley. Hold a single handle in the hand closest to the stack, with that arm reaching slightly behind you. Curl the handle up toward your shoulder while the elbow stays put.

The name is gym slang. It has nothing to do with Bayesian statistics. You will also see it called a facing away cable curl or a behind-the-body cable curl.

Two things are happening at once:

• Shoulder slightly extended. With the upper arm behind the torso, the biceps is lengthened across the shoulder as well as the elbow.
• A cable line of pull that never goes slack. The weight stays attached to your hand through the full path.

For a direct comparison with the preacher version, see our breakdown of preacher curls vs Bayesian cable curls.

Why does constant tension matter for a curl?

A standing dumbbell curl has a weak spot. Gravity pulls straight down, so when your arm hangs by your side the forearm is vertical and the dumbbell puts almost no turning force on the elbow. The biceps does little work in the first part of the rep, the load peaks around halfway, then eases off near the top.

A cable changes the direction of the pull. With the pulley low and behind you, the cable runs roughly horizontally back to the stack, which is close to perpendicular to your hanging forearm. That gives the biceps something to work against right at the start.

Why care about the stretched start? Mechanical tension is considered one of the main drivers of muscle growth, alongside muscle damage and metabolic stress, according to a widely cited review (Schoenfeld, 2010). And a growing set of studies suggests that training in the stretched position is a good place to put that tension, which we cover in our guide on stretch-mediated hypertrophy and range of motion.

How to do a Bayesian cable curl: step-by-step form

1. Set the pulley low. Wrist to hip height is the usual starting point. Clip on a single D-handle. If you are unsure about heights and angles, our cable setup guide for pulley heights and angles covers the basics.
2. Pick your stance. Turn away from the stack and stand in a staggered stance, with the foot on the working side behind. This stops the cable from pulling you backward.
3. Take a step forward. Walk out until the working arm sits slightly behind your torso and you feel a mild stretch in the biceps. More distance means more stretch, but the point is a controlled start, not the farthest reach.
4. Pin the elbow. The upper arm stays where it is. Let the elbow travel forward only a little as you curl. If the elbow swings forward a lot, the exercise drifts toward a standard curl.
5. Curl, squeeze, lower slowly. Take 1-2 seconds up and 2-3 seconds down. Let the arm fully lengthen at the bottom without letting the stack crash.
6. Reset between reps. Do not bounce at the bottom. The stretch is the whole point.

Cable friction and how to read the weight

Pulleys and cables are not frictionless, and each stack brand feels different. The weight you can curl on one machine will not match the next. Judge sets by effort, not by the number on the stack. Stop 1-3 reps short of failure and log the reps.

What does the research actually show?

Here is the honest part. No study I could find has tested the Bayesian curl directly. Everything below is related evidence, so the Bayesian-specific claim is an extrapolation.

Shoulder position changes which muscle grows

A 2026 trial in 21 untrained adults trained one arm with incline dumbbell curls (shoulder extended) and the other with preacher curls (shoulder flexed) for 12 weeks, using MRI to measure muscle volume. The biceps brachii grew more after incline curls (+22.1% vs +18.9%), while the brachialis (+17.1% vs +21.1%) and brachioradialis (+13.5% vs +17.6%) grew more after preacher curls. Total elbow flexor growth did not differ significantly (Kobayashi et al., 2026).

What this suggests: putting the arm behind the body (as in a Bayesian curl) nudges work toward the biceps. Caveats: untrained participants, dumbbells rather than cables, and differences described by the authors as small to moderate. For the full dumbbell version, see our guide to incline dumbbell curls and shoulder hyperextension.

Do stretched, long-range reps beat short ones?

The evidence is mixed, and that matters.

• A 2020 systematic review found that full range of motion helped lower-body hypertrophy, but evidence for the upper limbs was limited and conflicting (Schoenfeld and Grgic, 2020).
• A 2021 meta-analysis agreed: full range of motion was better for strength and lower-limb hypertrophy, with no significant superiority for muscle thickness overall (Pallarés et al., 2021).
• A 2025 trial in 30 resistance-trained people (within-person, 8 weeks) found lengthened partials and full range of motion gave similar elbow flexor and extensor growth. The authors suggested emphasizing the stretched position, by full range or lengthened partials (Wolf et al., 2025).
• A 2025 meta-analysis of 12 studies in young adults found training at longer vs shorter muscle lengths gave similar regional hypertrophy, with the authors urging caution because the length differences were small (Varovic et al., 2025).

Why might these disagree? Studies differ in muscle, exercise, how far “long” really is, and whether participants were trained. A plausible reading: the stretched position is a good place to train, but it is not magic, and short-term studies cannot tell us whether the Bayesian curl beats a well-done preacher or incline curl.

Bayesian curl vs other biceps curls

Exercise	Where it loads hardest	Shoulder position	Good for
Standing dumbbell curl	Mid-range	Neutral	Simple, easy to load
Incline dumbbell curl	Stretched	Extended	Long-head bias, dumbbells only
Bayesian cable curl	Stretched, easing near the top	Slightly extended	Tension at the bottom, one arm at a time
Preacher curl	Mid to shortened	Flexed	Strict form, brachialis and brachioradialis

You do not have to pick a winner. If you want more on free weights versus cables, see free weights vs cables vs machines for hypertrophy.

How many sets and reps should you do?

Start with 2-4 hard sets per arm, 8-15 reps, once or twice a week. Take the sets 1-3 reps from failure, with a controlled lowering phase.

More is not always better. A 2026 meta-regression of 67 studies found that gains in muscle size rise with weekly set volume, with diminishing returns (Pelland et al., 2026). Count sets from rows and pulldowns too, since they train the biceps indirectly. Our training volume calculator can help you add up weekly sets, and the free pull day plan shows where biceps work normally fits.

If you like to push past failure with partials in the stretched range, read the lengthened partials guide first. It is an advanced tool, not a starting point.

Common mistakes

• Cranking the stretch. More shoulder extension is not better. A mild biceps stretch is enough.
• Letting the elbow drift. If the elbow rides far forward, you lose the stretched start.
• Dropping the weight at the bottom. Control the lowering phase. The stretch under load is the point.
• Using the same weight on both arms. Most people have a weaker side. Match reps to the weaker arm.
• Treating it as your only curl. The 2026 trial above suggests the other elbow flexors respond a little more to shoulder-flexed curls, so rotating exercises is reasonable.

Is it safe? When to see a doctor

For most healthy lifters, a controlled Bayesian curl with a modest stretch is no more demanding than other curls. The arm-behind-body position does load the front of the shoulder and the long head of the biceps tendon, so ease in.

Back off if you feel sharp pain at the front of the shoulder, a pinch or pop in the biceps tendon, or elbow pain that lasts into the next day. See a doctor or physiotherapist if pain lingers, if you have weakness lifting your arm, or if you notice a new bulge or bruising in the upper arm. Research on this specific exercise in people with shoulder problems does not exist, so this is general caution, not a diagnosis. If you have tendon trouble, our guide to distal biceps tendonitis and elbow rehab is a good next read.

Doing it on a budget, Indian gym edition

Many Indian gyms have a single cable tower or a crowded functional trainer. You can still do this.

• One low pulley is enough. You do not need a fancy stack.
• Use a single D-handle. If there is only a rope or a bar, ask a trainer for a handle, or use the rope loop in one hand.
• No cable at all? A resistance band anchored at a low point behind you works. Band tension rises as it stretches, so it feels different, but the stretched start still applies.
• Hot, crowded gym? Do them early in the session while the cable is free, and keep a towel for grip.