Testosterone (T) administration to men increases lean body mass and decreases fat mass. Adiponectin is produced by adipocytes and is thought to influence insulin sensitivity. In this study, we sought to determine whether experimental alterations in serum T change adiponectin levels in normal men. We measured adiponectin levels in 28 healthy men ages 18-35 years before and during treatment with a potent gonadotropin-releasing-hormone (GnRH) antagonist, acyline. Decreased T levels led to increased serum adiponectin within 7 days; maximal adiponectin levels were reached on day 21 (baseline 8.6 +/- 0.9 compared with 12.2 +/- 1.0 microg/mL on day 21, P <.05) and persisted through day 30, despite no significant changes in body mass index (BMI) and an increase in leptin. The addition of T to acyline, maintaining serum T levels within the normal range, prevented the increase in adiponectin following acyline alone. In a second study, 25 men aged 55-85 years were treated with 3 weeks of high-dose T (testosterone enanthate [TE], 600 mg/wk intramuscularly). With high serum T levels, adiponectin levels decreased significantly by day 21 of treatment (baseline 14.3 +/- 1.9 compared with 10.8 +/- 1.5 microg/mL, P <.05 vs baseline and placebo), BMI slightly increased, and leptin levels were decreased. We conclude that adiponectin levels increase within days of experimental T deficiency in normal men, and the increase in adiponectin is prevented by T replacement. Furthermore, supraphysiologic T administration results in decreased adiponectin levels. Our data support the hypothesis that T, its metabolites, or both directly suppress adipocyte production of adiponectin.