
Also works: Traps
Lying with a foam roller under the upper back, arms crossed to spread the shoulder blades, you lift your hips and roll the area between the spine and the scapula, pausing 10-30 seconds on tender spots. Crossing the arms is what makes it work: it exposes the rhomboids that the shoulder blades otherwise cover completely.
Set up first, then execute. Every cue below is something you can check on your next set.
Breathing & tempo — Slow and continuous, roughly four seconds in and six out. When you park on a tender point, exhale into it — if you are holding your breath, you are pressing too hard.
Where to start, and what to change when it stops being hard.
Load is how much of your bodyweight you transfer onto the roller — regulate it by pushing more through your feet, or by keeping your hips on the floor for a lighter version. One to two minutes total for the whole upper back, up to daily. If a spot is still tender after two minutes, it needs training or rest, not more rolling.
The errors that show up most often, and the cue that fixes each one.
Cross your arms and hug your own shoulders hard. Without protraction the shoulder blades shield the rhomboids and you spend the whole set rolling scapula, not muscle.
Stop at the bottom of your ribcage. The lumbar spine has no ribs to protect it, so a roller there just extends unsupported vertebrae — take that region on the floor with knees hugged instead.
Support your head with your hands or keep the chin lightly tucked. Ten passes with a hyperextended neck will hurt more the next day than whatever you were rolling for.
Cap any single spot at 30 seconds and reduce your weight through the roller by keeping more pressure on your feet. Pain hard enough to hold your breath drives the guarding you are trying to remove.
With your arms at your sides, the shoulder blades sit flat over the rhomboids and the roller never reaches them — you just roll bone. Wrapping your arms across your chest protracts the scapulae around the ribcage and opens a strip of muscle either side of the spine that is otherwise inaccessible without another person's thumbs.
What foam rolling actually does is worth stating plainly: it produces a short-term drop in perceived tightness and a temporary increase in range, most likely through the nervous system rather than by changing tissue length. It does not break up adhesions or lengthen fascia. The effect is real, useful and lasts minutes to hours, not weeks.
One to two minutes before pressing or pulling work, when a temporary range increase is exactly what you want, or afterwards to take the edge off a stiff upper back. Because the effect is short-lived, the fix for a chronically tight mid-back is rowing volume and getting off the chair — the roller is symptom management.
No kit, too hard, or too easy — here's where to go instead.
No. The forces involved are nowhere near what it would take to deform dense connective tissue, and if they were, you would not be able to tolerate them. The measured effects are a short-term increase in range and a drop in perceived soreness, both most plausibly neural. Useful, but not what the marketing claims.
Minutes to a couple of hours for range of motion. That is why the sensible placement is immediately before the training or desk stretch you want the range for, rather than as a standalone ritual hours earlier.
It should be uncomfortable enough to notice and no more — around a 5 or 6 out of 10. If you are tensing up and holding your breath, you have gone past the useful range and are provoking exactly the protective tension you came to reduce.
Yes, and for the rhomboids it is often better because it fits into the strip between spine and shoulder blade that a wide roller bridges over. Stand with the ball between your back and a wall for control, and expect it to be considerably more intense than the roller.