In healthy middle-aged men, raloxifene treatment was associated with increased serum estradiol and decreased biochemical markers of bone turnover in subjects with estradiol levels below a threshold of 101.8 pM. INTRODUCTION: We investigated the effects of the selective estrogen receptor modulator raloxifene on bone remodeling in healthy middle-aged men. MATERIALS AND METHODS: Forty-three healthy eugonadal men (mean age, 56 years; range, 49-70 years) were enrolled in a randomized placebo-controlled, double-blind, two-sequence crossover study. The subjects received either raloxifene 120 mg/day or placebo for 6 weeks, followed by a 2-month washout period, before crossing over. To predict changes of urinary total deoxypyridinoline/creatinine on raloxifene treatment, we used a logistic regression model to determine cut-off values of sex hormones for highest sensitivity and specificity. RESULTS: In the whole group, raloxifene treatment was associated with an increase in serum sex hormones, that is, total testosterone (+13%, p < 0.01), bioavailable testosterone (+11%, p = 0.02), total estradiol (+11%, p < 0.002), and bioavailable estradiol (+11%, p = 0.035), and with a decrease in serum osteocalcin (-13%, p < 0.05) and serum total alkaline phosphatase (-6%, p < 0.05). Other biochemical markers of bone turnover remained unchanged. Using a logistic regression model to predict changes in urinary deoxypyridoline, we calculated thresholds for total (101.8 pM) and bioavailable (4.79 pM) estradiol, as well as for total (19.4 nM) and bioavailable (0.35 nM) testosterone. Raloxifene treatment was associated with an increase in serum estradiol and decrease in biochemical markers of bone turnover in men with estradiol values below these estradiol thresholds, without any significant change in subjects with values above them. Similarly, raloxifene treatment was associated with an increase in serum testosterone and a decrease in biochemical markers of bone turnover in those with baseline testosterone values below the testosterone thresholds. The association between antiresorptive effects of raloxifene and low sex hormone levels was more pronounced for estradiol than for testosterone. CONCLUSIONS: The antiresorptive effect of raloxifene was only detectable in men with low baseline estradiol levels. Unlike in postmenopausal women, the increase of estradiol may contribute to the antiresorptive effect of raloxifene in men.