Fifty-three acute or subacuta patients with low back pain were given standardized but optlmized activating conventional treatment by primary health care teams. Forty-eight patients received an experimental treatment that included specific manual treatment, such as manipulation and specific mobilization, muscle stretching, auto-traction, and cortisone lnjections. After 4 months, the experimental group had a less restricted range of movement in extension, loss restricted side-bending to the right and to the left, less local pain caused by extension and side-bending to the right, less pain radiating to the right leg caused by side-bending to the left, and a less positive straght leg raing test (both sides) than the conventionally treated group. Manual treatment was superior to the conventional activating treatment in normalizing pathologic findings on physical examination of the lower back. These results agree with the positive influence on pain, drug consumption, sick-leave, disability rating, and quality of life reported in other reports from the same study.