To determine an optimal dosage combination for a nylestriol=levonorgestrel (NYL/LNG) regimen in the treatment of female Sprague-Dawley (SD) rats with retinoic acid (RA)-induced osteoporosis in order to examine the rationale of NYL/LNG use for postmenopausal women. Methods. This study included two sets of experiments; one was designed to confirm the success of the RA-induced osteoporosis model involving 48 SD rats, and the other to determine the optimal dosage combination of NYL/LNG. In the second set of experiments, a total of 160 female SD rats at 7 months of age were randomly divided according to their basal body weights into 16 groups (10 in each). Treatment dosages of NYL and LNG were arrayed in a 2 factor-4 level (24) factorial design, i.e., NYL level I (N1 0.015mg/kg), level II (N2 0.05mg/kg), level III (N3 0.15mg/kg) and level IV (N4 0.5mg/kg) and LNG level I (L1 0.005mg/kg), level II (L2 0.015mg/kg), level III (L3 0.05mg/kg) and level IV (L4 0.15mg/kg). For 14 d, 70mg/kg=d RA was given intragastrically to establish the osteoporotic model and NYL and LNG were then administered in different dosages on alternate days over a 3wk period. Subsequently, body weight, uterine weight, endometrial status, Bone mineral density (BMD), bone turnover, and morphometry as well as parameters of biomechanics, serum sex hormones and lipids and estrogen receptor-a expression were determined for these rats. Results. Uterine weights at L2, L3, and L4 dosage levels were significantly lower than those at L1 (P<0.05). Serum estradiol at N4 and progesterone at N2, N3, and N4 had decreased. Bone mineral density at distal tibiae (R5) in L3 and L4, and at proximal femora (R3) in L4 had increased. The peak loading of lumbar vertebrae at N3 was higher than that at N1, N2, and N4 dosage levels and that of femora higher at N3 than that at N1 and N4. Serum alkaline phosphatase (ALP) levels at N3, N4, L2, L3, and L4 dosage levels were lower than those at N1, N2, and L1 (P<0.05). Moreover, there were lower levels of trabecular separation (Tb.Sp) at N3 and N4. The mineral apposition rates (MAR) at N2, N3, N4, L2, L3, and L4 were lower than those at N1 or L1. Bone formation rates (BFR) at L3 and L4 had significantly decreased as compared with that at L1. Levels of serum total cholesterol (TC) at N2, N3, and N4 were lower than that at N1 (P<0.05). Marked squamatization of uterine endometrium with an increased expression of estrogen receptor (ER)-a was observed at N4. Conclusion. The dosage of 0.15mg/kg NYL prevented bone loss, decreased bone turnover rate and increased the maximal loading of bone without obvious side effects in RA-induced osteoporotic rats. The addition of 0.015-0.15mg/kg of LNG (L2-L4) reduced uterine weights and serum ALP levels. Overall results showed that 0.15mg/kg of NYL in combination with 0.015mg/kg of LNG produced beneficial effects on bone metabolism in RA-induced osteoporotic rats.