This study was undertaken to determine whether the exercise-induced sparing of glucocorticoid-induced muscle atrophy is related to increased androgen cytosol binding. Female rats were divided into a sedentary or an exercise group that was trained by treadmill running 100 min/day for 13-15 wk. During the last 12 days of training, each of these groups was further subdivided into groups that received daily subcutaneous injections of cortisone acetate (CA) (100 mg/kg body wt) or the vehicle 1% carboxymethyl cellulose. Exercise prevented 30-40% of the weight loss due to CA treatment in gastrocnemius and plantaris muscles. Scatchard analyses of specific binding of [3H]methyltrienolone (R1881), a synthetic androgen that binds to androgen receptors, were nonlinear in muscles from vehicle-treated sedentary and trained rats and were resolved by a two-component binding model. The lower affinity component, which was attributed to a glucocorticoid receptor, disappeared in muscles of glucocorticoid-treated animals as evidenced by linear Scatchard plots. Receptor concentrations of the androgenic component of [3H]methyltrienolone binding were similar in gastrocnemius and plantaris muscles in all treatment groups. In binding specificity studies of gastrocnemius muscles, the relatively high competition by various glucocorticoids and progesterone for [3H]methyltrienolone binding in the vehicle-treated groups was reduced by CA treatment. The lack of change in androgen cytosol receptor levels suggests that this is not a mechanism by which exercise protects against glucocorticoid-induced muscle atrophy.