Can corrective exercise fix forward head and rounded shoulders?

Short answer

The angles do change. Corrective exercise improved head angle by about 3 degrees, and it was strength work, not stretching, that did it. But the idea that posture causes neck pain did not hold up in a study that followed people for 5 years.

Can corrective exercise fix forward head and rounded shoulders?
'Corrective exercise → straighter posture → neck and shoulder pain gone' — true?
It is often said that forward head and rounded shoulders must be fixed with exercise, and that once fixed, neck and shoulder pain goes away. We checked separately whether posture actually changes and whether posture and pain are really linked.
Exercise did change posture angles
901 people: all three angles improved
An analysis that pooled 28 randomized trials of corrective exercise (901 people) found large improvements in how far the head juts forward, how far the shoulders roll in, and how rounded the upper back is. This was compared with people who did not exercise.
By how much?
About 3 degrees of head angle
A 2026 analysis of 10 randomized trials (550 people) found the angle of the line from the ear to the base of the neck (how far forward the head sits) improved by 3.38 degrees more than in people who did not exercise.
Strength work changed it, not stretching
Stretching alone left posture unchanged
An analysis of 23 studies (969 people) found stretching, whether done once or for weeks, did not change posture. Strength training changed neck and upper-back posture a lot, but not lower back or pelvic posture.
So does posture cause pain?
Adults with neck pain had their head further forward
An analysis of 15 studies found adults with neck pain held their head further forward than adults without it, and the further forward, the worse the pain. Around age 17 there was no such difference. All were snapshots measured at one time, so they do not show which came first.
Followed over time, posture did not predict pain
Posture at 17, neck pain at 22
When 1,108 people were photographed at 17 and their sitting posture sorted into 4 types, neck pain and headaches did not differ between types. Following 686 of them to 22, posture had no link with lasting neck pain in men, and in women the slouched group actually had about 1/4 the risk of the upright group.
Pain dropped in people who already had neck pain
The change was less clear than the angle change
In 9 trials of only people with neck pain (480 people), pain scores fell by 1.42 points after corrective exercise. In the 901-person analysis of posture problems in general, changes in pain, function and strength did not show up as clearly as the angle changes.
Summary: angles change, the pain formula did not hold
About 3 degrees of head angle with strength work
Corrective exercise, especially strength training, changed forward head and rounded shoulder angles. But the idea that posture causes neck pain did not hold in a study that followed people for 5 years. Read corrective exercise as a way to change posture angles, without expecting pain to always vanish with it.