The putative skeletal effects of dietary soy phytoestrogens (SPE) were examined in comparison with those of conjugated equine estrogens (CEE; Premarin) in a 3-yr longitudinal study in ovariectomized female monkeys. Controls received alcohol-extracted soy protein with low phytoestrogen content, and treatment groups received either CEE (admixed into the control diet) or unextracted soy protein isolate containing SPE. The acknowledged bone protective effect of CEE was reflected by higher bone mass (by dual energy x-ray absorptiometry) and lower bone turnover marker levels. In contrast, control and SPE groups lost significant lumbar spine bone mineral content and density and whole body bone mineral content within the first year, resulting in reduced bone mass for both groups compared with CEE (P < 0.0005). No effect of SPE was observed for any bone mass measure (P > 0.44), although transient, estrogen-like effects of SPE on serum alkaline phosphatase, calcium, and C-terminal cross-link of type I collagen were observed at 3 months (P < 0.02). These results suggest that SPE may be poor substitutes for mammalian estrogens in protecting against bone loss resulting from estrogen deficiency.