The purpose of this study was to examine the influence of estrogen-progestin replacement therapy and exercise on the lumbar spine mobility and back symptoms of early postmenopausal women. The population sample consisted of 78 healthy, 49- to 55-year-old women, 0.5-5 years after menopause, who were randomized into three groups, two receiving different protocols of estradiol valerate combined with medroxyprogesterone acetate replacement therapy, and the third group a placebo. These groups were then randomized into exercise and control cases and monitored for 2 years. The mobility of the lumbar spine was measured and symptoms investigated using the Million and Oswestry pain and disability questionnaires and pain drawings at the baseline and after 1 and 2 years. During the follow-up, the mobility of the lumbar spine decreased in all six groups. The decrease was most evident in those who had been the most flexible at baseline (P < 0.0001). The decrease was less notable in the hormone replacement therapy groups than in the control group. When the replacement therapy groups were pooled together, the difference was significant at a P < 0.05 level. No difference was seen between the hormone combinations. The exercise intervention was insufficient to influence lumbar spine mobility. Only sporadic cases of back symptoms appeared and disappeared among the subjects during the follow-up, and no preventive or aggravating effects of hormone replacement therapy or the exercise program on symptoms were detected.