To extend and confirm previous data, we examined the effects of raloxifene on the proximal tibia of ovariectomized rats, aged 6 months, longitudinally and cross-sectionally by computed tomography (pQCT) and then compared the effects to those of orally dosed estrogen. Comparative analysis of phantoms and rat bones showed that the pQCT is precise and correlates with a Hologic QDR 1000W (DXA) with R = 0.999 but is capable of measuring significant differences between groups when the DXA cannot. This may reflect the ability of the pQCT to determine bone volume, mineral content (mg) and volumetric mineral density (mg/cm3), compared with two-dimensional analyses performed with DXA. Longitudinal analysis of the proximal tibia in vivo showed a significant 17% reduction in mineral density 31 days after ovariectomy. Examination of the images from ovariectomized rats showed a progressive increase in the cross-sectional area of the proximal tibiae, loss of trabecular bone, widening of marrow spaces and thinning of the cortical bone wall opposite the fibula. Regression analysis of the dose-dependent protective effects of raloxifene showed the half-maximal efficacy on tibiae mineral density to be ED50 = 0.4 mg/kg/day per os by pQCT and 0.2 mg/kg/day by DXA. By comparison, 17 alpha ethynyl estradiol showed dose-dependent effects with ED50 = 0.013 mg/kg/day per os by pQCT. Both raloxifene and ethynyl estradiol had beneficial effects on serum lipids, producing 50% reduction of cholesterol at 0.1 mg/kg/day raloxifene and 80% reduction with 0.01 mg/kg/day ethynyl estradiol. However, raloxifene up to 10 mg/kg/day had little effect on uterine weight, whereas 0.01 mg/kg/day ethynyl estradiol increased uterine wet weight by 300%.(ABSTRACT TRUNCATED AT 250 WORDS)