Hepatic cirrhosis is frequently listed as a cause of gynecomastia. We found previously that in hospitalized men the prevalence of gynecomastia was correlated with body mass index and with age. The mean body mass index and the prevalence of gynecomastia in the cirrhotic subjects (nonedematous) did not differ from those in the overall population. Because more severely cirrhotic subjects with ascites, peripheral edema, or both usually are thin, we examined 18 more severely cirrhotic subjects and 18 nonobese (mean body mass index, 20.9 +/- 0.6 kg/m2), age-matched control subjects for the prevalence of palpable gynecomastia. Total testosterone, free testosterone, total estrogen, and estradiol concentrations also were measured. Fifty percent of the control subjects had gynecomastia. Breast tissue diameter was correlated with body mass index. The prevalence of gynecomastia in the cirrhotic subjects was 44%. In these subjects no significant correlation was noted between breast tissue diameter and body mass index, presumably because the body mass index was increased owing to fluid retention. The results could not be accounted for based on medications. Serum free testosterone concentrations were lower in the cirrhotic patients than in the controls (0.11 +/- 0.02 vs 0.22 +/- 0.03 nmol/L). The total estrogen-free testosterone ratio was higher in cirrhotic patients (10.3 +/- 2.5 vs 2.6 +/- 0.5), as was the estradiol-free testosterone ratio (2.2 +/- 0.7 vs 0.5 +/- 0.1). These ratios did not differ significantly in cirrhotic subjects with and without gynecomastia. Therefore, these data indicate that factors other than the estrogen-testosterone ratio are playing a role in the development of gynecomastia in both cirrhotic subjects and controls or that breast tissue sensitivity to an elevated estrogen-testosterone ratio is highly varible.