Rows and pulling movements: the unshowy half of training
Pulling movements get less attention than pressing ones, although they deserve the same volume. This text sorts the families — horizontal and vertical pulling — and describes what actually matters in execution.
Two directions
The back is not a single muscle group but a surface of several layers with differing fibre directions. It therefore cannot be covered by one movement. Pulling movements can be sorted roughly by direction. Horizontal pulls — the various rowing forms — bring the load towards the torso. Vertical pulls — pull-up, lat pulldown — bring it from above downwards.
Both work the broad back muscle but set different emphases. Horizontal pulling additionally loads the musculature between the shoulder blades and the rear portion of the shoulder more. Vertical pulling emphasises the broad back muscle across a larger range.
A plan containing both directions covers the back more completely than one restricted to a single direction. That is not a rule from research but a plain consequence of which muscles work in which direction.
The most common rowing forms
In the bent-over barbell row the torso is inclined and the load is pulled to the abdomen. This form additionally works the back extensors and the trunk, because the position has to be held against gravity — an advantage and at the same time the reason this form gives way first under fatigue.
In the supported row — one hand and one knee on a bench, dumbbell in the other hand — that share largely disappears. The same musculature can therefore be loaded cleanly for longer, and both sides work separately.
On a seated cable row, resistance stays more even across the whole movement than with free weights, because gravity does not act vertically. That is the practical reason to choose this variant for additional volume.
What counts in execution
The most common error is not posture but sequence. Starting the movement with the arms lets the arm flexor take over much of the work and fatigue before the back is fully worked. More sensible is to set the shoulder blades moving first and let the arms follow.
The second point is range at the end. Bringing the load only halfway shortens exactly the section where the musculature between the shoulder blades works. Conversely, yanking the load to the body with momentum from the trunk is no gain.
In the bent-over row, the torso drifts upwards with increasing fatigue without being noticed. A fixed reference point — the torso angle, the height at which the bar meets the body — makes that drift visible.
Grip and hand position
Whether the grip is palms down, palms up or facing each other changes the involvement of the arm flexor and the angle at which the shoulder works. The differences in back involvement are smaller in studies than the debate suggests.
More relevant in practice is that grip strength gives out first in many pulling movements. Straps can play the same role here as in the deadlift: they shift the limiting factor when that is intended.
The rear shoulder and the small muscles
Alongside the large back muscles, several smaller groups work in pulling movements: the rear portion of the shoulder, the muscles drawing the shoulder blade towards the ribcage, and the external rotators of the shoulder.
These groups are loaded during heavy pulling but not always to their limit — under high load the large muscles take the main share. Addressing them specifically usually means lighter movements with a larger range, such as pulls on a high cable or lateral raises in a bent-over position.
Whether this additional work brings a measurable advantage beyond general pulling is less clear than the popularity of such exercises suggests. As a low-effort complement it is unobjectionable; as a mandatory programme it cannot be justified.
How much pulling relative to pressing?
Ratios circulate in training practice — equal amounts, or more pulling than pressing. They rest on the observation that pressing movements are more common anyway in everyday life and in many plans, and on the assumption that balancing helps the shoulder.
The second assumption is plausible but not well established; controlled studies are scarce. What is evident without a study: a plan with similar volume in both directions covers more musculature and costs nothing. More than that cannot be derived from the evidence — and planning needs no more.
The key points
- Horizontal and vertical pulling set different emphases; both belong in a plan.
- The bent-over row additionally loads trunk position — an advantage and the place that gives way under fatigue.
- The movement starts with the shoulder blades, not the arms.
- A cut-short end range removes exactly the section where the musculature between the shoulder blades works.
- Differences between grip types are smaller in studies than the debate about them.
- Similar volume in pulling and pressing covers more musculature; claims beyond that are not supported.
Last reviewed: 14.08.2026
Created by AI