The present study has been undertaken to demonstrate the effect of 17β-estradiol on the healing of tibial bones after marrow ablation. Ninety-six 4-month-old female rats were divided into three experimental groups: group 1 was subjected to ablation in both tibiae and to injection of vehicle; group 2 to ablation in both tibiae and estradiol administration, and group 3 to estradiol administration without ablation. Before the rats were killed, all on the same day, 8 animals of each group underwent treatment for 3, 6, 14 and 21 days, respectively: every second day, 0.1 ml arachis oil was injected intramuscularly into group 1 and 17β-estradiol into groups 2 and 3. An additional 8 untreated animals were used as controls. Tibial bones were studied chemically and morphologically. While the control and ablated animals gained weight, there was a significant decrease in the gain in body weight of estradiol-treated rats. Bone and ash weight were increased in all experimental groups. The ratios (%) of tibial ash weight and of tibial Ca and Mg contents to body weight significantly increased in all experimental groups, compared to the controls; whereas P increased only at 6 and 14 days. As shown by computerized histomorphometry, the height of the proximal tibial growth plate was increased following ablation, but not with estradiol treatment. The proportions by area of calcified cartilage and metaphyseal trabeclar bone were significantly lower and the percentage of bone marrow was higher in all groups treated for 3 and 6 days; the values were similar to those of the controls with 21 days of treatment. The results of this study indicate that intramuscular administration of estradiol to adult rats, after marrow ablation, did not significantly influence bone healing, as observed following ablation alone. It seems that the effects of estradiol on the bone of adult rats with slowly remodelling bones does not change with the increase in bone turnover, and that estradiol has little effect on endosteal bone undergoing enhanced remodelling.