In a randomized, observer-blind trial, 150 men and women, aged 21-64 years, with chronic/subchronic low-back pain, followed one of these three treatment regimens: 1) intensive, dynamic back-muscle exercises; 2) conventional physiotherapy, including isometric exercises for the trunk and leg muscles; and 3) placebo-control treatment involving semihot packs and light traction. Eight treatment sessions were given during the course of 4 weeks, each session lasting 1 hour. The short-term effect was evaluated at the end of the treatment period and 1 month later, and the long-term effect at 6 and 12 months. The evaluations included recording of changes in pain level and assessment of overall treatment effect, which were indicated on visual interval scales. Subgroups of patients could be identified according to their treatment responses: physiotherapy was the superior treatment for the male participants, whereas the intensive back exercises appeared to be most efficient for the female participants. Patients with moderate or hard physical occupations tended toward a better response with physiotherapy, whereas intensive back exercises seemed most effective for those with sedentary/light job functions.