Smith Machine RDL: How to Set Up, Load, and Lower It

The Romanian deadlift is the cleanest way most lifters have to load the hamstrings and glutes through a hip hinge. Doing it on a Smith machine is popular because the bar cannot wander, and that stability is real.

It is also the exact reason a Smith machine RDL is a compromised hinge, and pretending otherwise leads to bad setups and sore lower backs. This guide covers what changes, what you have to do about it, and when to walk over to the dumbbell rack instead.

What a Romanian Deadlift Is, and What the Smith Machine Changes

A Romanian deadlift starts at the top, from a standing position with the bar already in your hands, and lowers under control until the hamstrings run out of length. The knees stay soft and mostly still, the hips travel back, and the torso pitches forward over them.

It is not a deadlift off the floor and it is not a stiff leg deadlift with locked knees. The whole movement lives in the hips.

On a barbell, three things move at once: the hips go back, the shoulders drift forward, and the bar stays roughly over the middle of your foot because the combined weight of you and the bar has to stay balanced there. Your body solves that balance problem automatically, rep after rep, without you thinking about it.

A Smith machine takes that problem away and replaces it with a different one. The rails carry any sideways force, so nothing has to balance, and the bar travels a fixed vertical or slightly angled line no matter what your hips are doing.

The Fixed Bar Path Is the Whole Problem

Here is the mechanical conflict in one sentence: a hinge wants the bar to sit over the mid foot while the hips travel back behind it, and a Smith rail refuses to let the bar move even a fraction of an inch to make that happen.

On a barbell RDL your shoulders travel forward as your hips travel back, and the bar hangs from your shoulders, so the bar keeps finding the balanced spot on its own. On a Smith machine your hands are locked to the rail, your arms hang straight down, and your shoulders are therefore pinned in place horizontally for the entire rep.

That means every bit of the movement your torso would normally split between forward shoulder travel and backward hip travel now has to come out of hip travel alone. Your hips have to go further back than they would on a barbell for the same amount of forward lean.

If you set your feet directly under the bar, the way you would with a barbell, that extra hip travel puts your weight behind your heels at the bottom, and the rail quietly holds you up. You are no longer standing on your own, and the lift stops teaching your hips anything about balance.

The fix is to set your feet forward of the bar line at setup, so the bar ends up sitting back toward your heels where a barbell would naturally have drifted, and so your own weight stays over your mid foot at the bottom of the rep.

How Far Forward to Set Your Feet on a Smith Machine RDL

Most guides say a few inches forward and stop there, which is not something you can actually set up with.

The offset you need is roughly the distance your shoulders would have travelled forward on a barbell RDL, because that is the travel the rail has taken away from you.

The formula, in plain text:

Foot offset forward of the bar, in inches = 0.35 x torso length in inches x sine of your torso pitch angle at the bottom

Worked example. A lifter with a 23 inch torso who reaches about 70 degrees of forward pitch gets 0.35 x 23 x 0.94, which is 7.6 inches, so call it 7 to 8 inches forward of the bar line.

Every assumption, stated out loud:

• Torso length is measured standing, from the hip crease to the top of the shoulder. For most adults between 5 foot 4 and 6 foot 2 that lands somewhere between 20 and 26 inches.
• Torso pitch is how far your trunk tips from vertical at the bottom of the rep. Most lifters finish somewhere between 60 and 75 degrees, and 70 is a reasonable default.
• The 0.35 is the share of total torso sweep that goes into forward shoulder travel on a free barbell RDL rather than backward hip travel. It is a rounded planning figure, and it is the softest input in the whole calculation.
• The rail is assumed to be vertical.

How the answer moves if an assumption moves. Pitch barely matters: dropping from 70 to 60 degrees only shifts the answer from 7.6 to 7.0 inches, because the sine curve is nearly flat up there.

The 0.35 share matters a lot. At 0.25 the same lifter gets 5.4 inches and at 0.45 they get 9.7 inches, so the honest output is a band of roughly 5 to 10 inches rather than a single number.

Torso length matters almost as much. A 20 inch torso gives 6.6 inches and a 26 inch torso gives 8.6 inches at the same pitch.

Angled rails move it again. If your machine tilts so the bar drifts toward you on the way down, you need less offset, and if it drifts away from you, you need more.

Treat it as a starting mark, not a rule: this is a WorkoutHealthy planning estimate built from segment geometry, not an industry standard.

Set your feet at the calculated distance, do one light set, and check two things at the bottom: your weight should feel spread across the whole foot, and the bar should be close to your legs without dragging on them. Move your feet an inch at a time until both are true.

Setting the Bar Height and the Start Position

Rack the bar so it sits against the front of your thighs when you stand tall, which for most lifters is mid thigh to just below the hip crease.

Too high and you have to bend your knees to unrack, which starts you in a squat. Too low and you are already hinged before the set begins, which wastes the top of the range.

Grip just outside your thighs, unrack, take your calculated step forward, and set your feet hip width with the toes straight ahead. Pull your shoulders down and back, brace, and let the knees soften a few degrees.

How Far to Lower, and Finding the Real End of Your Range

Your depth is set by your hamstrings, not by the bar reaching your shins and not by what the person next to you is doing.

The end of the range is the exact moment your lower back starts to round, and one inch past that point you are no longer training hamstrings, you are loading a flexed spine. Stop at the last rep depth you can hold with a flat back, and treat that as your working depth for the whole set.

Two ways to find it. Have someone film one set from the side and watch for the first frame where the lumbar curve reverses, or run your own hand along your lower back on a light set and feel for the point where the arch disappears.

Most lifters land between mid shin and just below the knee, and plenty of people with long femurs or tight hamstrings stop at the top of the kneecap. That is a normal finishing depth, not a failure.

The same logic governs depth on every big lower body lift, which our guide to finding your honest squat depth works through in detail.

Hinge or Squat? How to Tell Which One You Are Doing

The most common way a Smith RDL goes wrong is that it quietly turns into a bad squat, and the lifter never notices because the bar path looks identical either way.

Use this check. Film one set from the side and watch your knee angle from the first rep to the bottom, because in a hinge the knee angle barely changes and in a squat it opens up by 40 degrees or more.

The second check is where you feel it the next day. A hinge leaves the hamstrings and glutes sore, and if your quads are sore instead, you squatted.

The third check is the shin. In a hinge your shins stay close to vertical for the whole rep, and if the knees travel forward over the toes you are squatting the weight up.

Sets, Reps, Tempo, and Load

The RDL responds well to controlled lowering, so the eccentric is where the programming effort should go rather than the lift back up.

A workable default is 3 to 4 sets of 8 to 10 reps, with a 3 second lowering phase, a brief pause at the bottom where you feel the hamstrings loaded, and a normal speed return to standing.

Rest 2 to 3 minutes between sets. Run it once or twice a week, and if you are running it twice, make the second session lighter and shorter rather than a repeat of the first.

On load, start lower than you think. In my view most lifters should begin a Smith RDL at roughly 60 percent of the load they use for a barbell RDL at the same rep count, then add weight only when all the reps hold the same depth and the same flat back.

The fixed path makes heavy weight feel more manageable than it is, because balance is no longer part of the difficulty. For how rep ranges and rest map to different goals, our strength programming guide covers the tradeoffs.

The Errors That Show Up Most

Bending the knees too much. The knees soften and then stop, so if they keep travelling you have started squatting, and the fix is usually to slow the descent and re cue the hips going straight back.

Rounding the lower back to chase depth. Depth is an output of your hamstring length, not a target you push toward, and going lower by giving up the arch trades the training effect for spinal load.

Going too heavy because the machine feels stable. This is the specific failure mode of the Smith version and it is worth naming on its own.

Yanking the bottom position. Reversing hard out of full stretch is where hamstring strains happen, so pause briefly and start the return smoothly.

When the Barbell or Dumbbell RDL Is Simply the Better Choice

Reach for the barbell version when you want the hinge to carry over to anything you do standing, because balance and bar control are part of the skill and the Smith removes both.

Reach for dumbbells when your hamstring length or your proportions make the fixed path awkward, since dumbbells let the load find its own line and give you a much easier single leg option.

Reach for either free weight version when your lower back is the limiting factor, because shifting the load an inch mid rep is a safety valve a rail does not give you. Our article on lower back dumbbell exercises is a better starting point than adding load here.

The Smith version earns its place in three situations: when you are training alone and want a fixed depth you can repeat exactly, when you are doing high rep hamstring work at the end of a session and balance is already fried, and when a shoulder or wrist issue makes holding a free bar in a fixed grip more comfortable than letting it move.

What the Research Actually Supports

The direct evidence on Smith machine RDLs specifically is thin, and what exists is adjacent rather than direct.

On the machine question, a 2009 study in the Journal of Strength and Conditioning Research compared free weight squats to Smith machine squats and found muscle activity averaged across the measured muscles was 43 percent higher in the free weight version, with the biceps femoris 26 percent higher (Schwanbeck, Chilibeck, and Binsted, 2009). That was the squat and not the RDL, so treat it as a directional hint about fixed bar paths rather than a measurement of this exercise.

On the movement itself, a systematic review of deadlift electromyography found the Romanian deadlift produced lower erector spinae activation relative to the hamstrings than the conventional deadlift did, which is the specific reason it is a useful hamstring lift rather than a back lift (Martin-Fuentes, Oliva-Lozano, and Muyor, 2020).

On the eccentric emphasis, a 2025 trial found that 6 weeks of eccentrically biased Romanian deadlifts increased biceps femoris long head fascicle length by about 9 percent and hamstring cross sectional area by about 10 percent, with no difference from the Nordic hamstring exercise (Crawford and colleagues, 2025). That study used a barbell, and the fascicle length gains disappeared after 2 weeks without training, which is a useful reminder that this adaptation is a use it or lose it one.

Romanian Deadlift Hinge Demonstration

The machine fixes the bar path, but the underlying hip-hinge mechanics are the same as the free-weight RDL shown here.

Where to Go Next

To see where the RDL sits against your other pulls, our comparison of rack pulls and deadlifts works through the same shortened range tradeoff.

References

1. Schwanbeck S, Chilibeck PD, Binsted G. A comparison of free weight squat to Smith machine squat using electromyography. Journal of Strength and Conditioning Research. 2009;23(9):2588 to 2591. https://pubmed.ncbi.nlm.nih.gov/19855308/
2. Martin-Fuentes I, Oliva-Lozano JM, Muyor JM. Electromyographic activity in deadlift exercise and its variants. A systematic review. PLOS ONE. 2020;15(2):e0229507. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0229507
3. Crawford SK, Sandberg C, Vlisides J, Thompson Q, Mosiman SJ, Heiderscheit BC, Hickey JT. Hamstrings muscle architecture and morphology following 6 wk of an eccentrically biased Romanian deadlift or Nordic hamstring exercise intervention. Medicine and Science in Sports and Exercise. 2025;57(8):1799 to 1809. https://pubmed.ncbi.nlm.nih.gov/40085810/