Is it okay to do strength training when your back hurts?

Short answer

Yes. For back pain lasting more than 3 months, strength training reduced pain, and heavy deadlifts were no worse than light exercises at reducing pain. But for pain that started less than 6 weeks ago, exercise did not reduce pain by a noticeable amount.

Is it okay to do strength training when your back hurts?
'Back pain → lifting makes it worse → rest' — does this formula hold?
A common belief says that when your back hurts, you should stop lifting and lie down. We pooled studies on common back pain with no specific underlying disease to see whether strength training makes pain worse or better.
Exercise worked for long-lasting back pain
Across 249 trials, pain dropped 15 points out of 100
An analysis of 249 trials on back pain lasting more than 3 months found that people who exercised had 15.2 points more pain relief (on a 100-point scale) than those who did nothing or continued usual care. That beats the 15-point threshold the researchers set in advance for a difference people can feel.
Strength training alone did the same
10 trials, 434 people: pain went down
A 2026 analysis that pooled only strength training trials (10 trials, 434 people) found less pain and less trouble with daily life in the strength training groups. The pain drop was large enough for people to feel. In the 249-trial analysis, side effects of exercise were mostly mild, like muscle soreness.
So what about deadlifts?
Heavy deadlifts were no worse at reducing pain
70 people with recurring back pain did 12 sessions over 8 weeks of either heavy deadlifts or light posture-control exercises. Pain went down in both groups, with no difference between them. Daily activity scores improved more with the light exercises, though: 4.2 points vs 2.5.
Is heavier weight better?
No — it was close to bodyweight exercise
An analysis of 13 trials (778 people) found that lifting machines or dumbbells reduced pain only 0.52 points more (on a 10-point scale) than bodyweight exercise, and only after 7 weeks. Other time points and daily-life trouble showed no difference. The researchers concluded that lifting weights is safe.
But new pain was different
For pain under 6 weeks old, exercise gave no noticeable effect
An analysis of 23 trials (2,674 people) on back pain that started less than 6 weeks earlier found that exercise did not reduce pain or daily-life trouble by a noticeable amount compared with fake treatment or no treatment. Exercise worked for long-lasting pain.
Exercise also prevented the next episode
Exercise plus education cut the risk of back pain by 45%
An analysis of 21 trials (30,850 people) found that exercise combined with education lowered the risk of a new back pain episode by 45%. Exercise alone lowered it by 35%. Back belts and education alone did not lower the risk.
Bottom line: strength training is fine for long-lasting back pain
The 'you must rest' formula didn't hold
For back pain lasting more than 3 months, strength training reduced pain, and heavy weights were no more harmful than bodyweight exercise and worked about the same. Starting with weights you can handle without too much pain and keeping at it works. New pain did not heal faster with exercise.