You've heard the reasoning enough times that it feels settled. RDLs are compound, they load the hamstrings with serious weight, and leg curls are the isolation version of the same target. If the compound covers the muscle group, the isolation is redundant.
Except the reasoning rests on one word doing too much work. 'Hamstrings.' Four separate muscles make up the group, each with different attachment points across the hip and knee joints. RDLs load your hamstrings — but which of the four, and which ones do they leave behind?
Do You Need Leg Curls if You Do RDLs?
Your hamstrings contain four individual muscles that respond differently depending on whether the exercise moves the hip or the knee. Hip-extension movements like RDLs primarily grow the long head and semimembranosus. Knee-flexion movements like leg curls primarily grow the short head and send a stronger signal to the semitendinosus. Neither pattern alone covers all four.
— Bourne et al. 2017 · Br J Sports Med · n=30
In a ten-week trial that measured each hamstring muscle individually on MRI, hip-extension training and knee-flexion training grew the four muscles in measurably different patterns. The long head of the biceps femoris — the muscle most lifters picture when they say 'hamstring' — grew significantly under hip-extension training. Under knee-flexion training alone, its growth didn't significantly clear the group that did nothing.
The semimembranosus, the deepest of the four, told the starkest story. After ten weeks of knee-flexion-only training, it grew no more than the group that never trained at all. Hip-extension training grew it. Knee-flexion training didn't register.
The reverse gap was just as real. The short head of the biceps femoris — the only hamstring that crosses the knee joint alone, without reaching the hip — grew significantly under knee-flexion training. Hip-extension training left it largely untouched. And the semitendinosus, the longest of the four, grew under both movement patterns with no significant difference between them.
| Muscle | RDLs | Leg Curls |
|---|---|---|
| Long head | Grew | Didn't clear control |
| Short head | Barely responded | Grew |
| Semimembranosus | Grew | Same as control |
| Semitendinosus | Grew | Grew |
Each of the four muscles has a different lever advantage depending on which joint is moving. The long head has a larger moment arm at the hip, which gives it more mechanical advantage during movements like the RDL. The semitendinosus has a larger moment arm at the knee, which is why it dominates during knee-flexion movements. The short head doesn't cross the hip at all, so hip-extension exercises can't load it directly.
A caveat the evidence demands. The exercises in the growth trial were a 45-degree hip extension machine and the Nordic hamstring exercise, not literally an RDL and a leg curl. Both sit in the same movement classes — hip-dominant and knee-dominant hamstring loading. The stiff-leg deadlift and the leg curl have both been measured directly in a broader activation analysis, and both fall on the expected sides of the divide. The inference is reasonable, but that gap between inference and direct test is worth naming. The participants were also recreationally active men, and whether these exact patterns hold across different training backgrounds hasn't been tested.
Your RDLs cover the long head and the semimembranosus. Your leg curls cover the short head and send the stronger signal to the semitendinosus. One muscle responds well to both. Neither movement alone reaches all four. The real question might not be whether you need leg curls. It might be whether per-muscle regional responses reshape how you structure the rest of your program.