OBJECTIVES To evaluate the value of peripheral quantitative computed tomography (pQCT) in measuring bone architecture and biomechanic properties. METHODS 50 virgin female Wistar rats six months old were randomly divided into 4 groups: (1) 8 rats were killed as baseline group; (2) 8 rats underwent sham operation and then were killed 14 weeks after (sham operation group); (3) 16 rats underwent bilateral ovariectomy (OVX) without further intervention. Six and 14 weeks after the operation each 8 rats were killed (OVX group); and (4) 18 rats underwent OVX too. After the OVX 9 of the 18 rats were treated with 17beta-estradiol 20 micro g/kg/d IH and 9 rats were treated with estradiol valerate 800 micro g/kg/d po for 8 weeks respectively. Then the 18 rats were killed (OVX plus estrogen group, O + E group). The right tibiae of the rats were taken for histomorphometric analysis, and the right femora were prepared for pQCT scanning and bone biomechanical measurement with indentation test and three-point bending test. RESULTS Histomorphometric analysis showed that the trabecular volume of proximal tibia (Cn-BV/TV) in the OVX group was 8.1 +/- 1.4%, significantly lower than that in the sham operation group (19.5 +/- 1.5%, P < 0.01). pQCT scanning showed that the femoral trabecular bone mineral content (Trab BMC) in the OVX group was 1.7 +/- 0.3 mg/mm, significantly lower than that in the sham operation group (3.2 +/- 0.5 mg/mm, P < 0.01) and the femoral trabecular bone mineral density (Trab BMD) in the OVX group was 158 +/- 32 mg/mm(3), significantly lower than that in the sham operation group (320 +/- 39 mg/mm(3), P < 0.01). The cancellous maximal load (Can load) of the distal shaft of femur in the OVX group was 12.5 +/- 2.5 N, significantly lower than that in the sham operation group (45.9 +/- 3.2 N, P < 0.01). The cancellous stiffness (Can Stiff) of the distal shaft of femur in the OVX group was 226 +/- 48 N/mm, significantly lower than that in the sham operation group (396 +/- 72 N/mm, P < 0.01). The Can load of O + E group was 21.8 +/- 3.7 N, significantly higher than that in the OVX group (P < 0.05). The Can Stiff of the O + E group was 382 +/- 54 N/mm, significantly higher than that in the OVX group (P < 0.05). There were no significant differences in cortical bone determined by pQCT as well as biomechanic properties in measured by three point test after OVX and estrogen treatment. A significant positive correlation was shown between Trab BMD and Cn-TV/BV and between Trab BMD and Tb N (r = 0.88 and 0.73, both P < 0.01). Similarly, both Trab BMC and Trab BMD of the femur were significantly correlated with the Can load and Can Stiff determined by indentation test (r = 0.47 - 0.68, all P < 0.01). There was also a significant correlation of parameters measured by pQCT in cortical bone with the maximal load and stiffness for the femur midshaft, and the best correlation was found between the maximal load of femur midshaft and Crt BMC and Crt A (both r = 0.76 and P < 0.01). CONCLUSION The geometric, densitometric and mechanical properties in cortical and trabecular bones of rat can be well described by pQCT.