100 women who had taken part in a prospective controlled trial of oestrogen therapy for prevention of post-oophorectomy bone loss were reviewed after a median follow-up period of nine years. A significant reduction in height occurred among the placebo-treated group, but not in the group treated with mestranol (mean 23·3 μg/day). The placebo-treated group had a higher spine score, lower central vertebral height, and larger wedge-angle than the oestrogen group. Within each group none of these spinal morphometric changes correlated with changes in mineral content of metacarpal or radial bones as measured by photon absorptiometry or X-ray densitometry, although both peripheral and central measurements showed highly significant differences between groups. Oestrogen treatment, therefore, prevents against central, as well as peripheral, bone loss, and reduces the incidence of vertebral compression.