The essential role of estrogen (E) in regulation of developing peak bone mass in males was confirmed when young adult men were described who cannot respond to or produce E because of defective E receptor alpha or P-450 aromatase enzyme, respectively. These men had significantly reduced bone mineral density (BMD) despite normal or supranormal androgen concentrations, and E administration improved BMD in the men with aromatase deficiency, whereas testosterone (T) was ineffective. Because new P450 aromatase inhibitors may prove to be potential drugs in various growth disorders, the effect of suppression of E action on developing peak bone mass has to be closely evaluated. In this study, we explored the effects of suppression of E synthesis on bone metabolism in pubertal boys. A total of 23 boys with constitutional delay of puberty were randomized to receive T and placebo or T and a specific and potent P450 aromatase inhibitor, letrozole. We determined BMD in the lumbar spine and the femoral neck. Bone resorption was studied by measuring the serum concentration of cross-linked carboxyterminal telopeptide of type I collagen by two different methods (CTx and ICTP), and bone formation by determining the serum concentrations of carboxyterminal propeptide of type I procollagen (PICP), osteocalcin, and alkaline phosphatase. We demonstrated previously that, during treatment with T and placebo, the concentrations of androgens and E increased. During treatment with T and letrozole, the E concentrations remained at the pretreatment level, but the androgen concentrations increased; the increase in the T concentration was more than 5-fold higher than during treatment with T and placebo. We did not observe any significant differences in the changes in bone mineral content, BMD, or bone mineral apparent density, an estimate of true volumetric BMD, between the treated groups. Lumbar spine bone mineral apparent density increased in both treated groups; but in the T- plus letrozole-treated group, the increase was statistically significant only 6 months after discontinuation of letrozole treatment. All bone resorption and formation markers increased during treatment with T and placebo. During treatment with T plus letrozole, CTx, PICP, and osteocalcin remained unchanged, whereas ICTP and alkaline phosphatase increased. Thus, 1-yr treatment with this new P450 aromatase inhibitor in pubertal boys is unlikely to be associated with any major harmful effect on developing peak bone mass. However, to convincingly exclude such effects, particularly rare or minor ones, will require a study with a larger sample size; and thus, close follow-up of bone metabolism during treatment with P450 aromatase inhibitors is still warranted.