The urinary levels of seven steroids, 5 alpha-androstane-3 alpha,17 beta-diol, 5 beta-androstane-3 alpha,17 beta-diol, androsterone, etiocholanolone, tetrahydrocortisone, tetrahydrocortisol and allotetrahydrocortisol were measured in both normal (n = 18) and hirsute (n = 24) women. The results confirmed 5 alpha-androstane-3 alpha,17 beta-diol as the most significant steroid with respect to discrimination between hirsute and normal subjects. Investigation of the inter-steroid relationships, using multivariate techniques established that the mode of steroid metabolism was different between the two groups. Whereas in normal women the strong correlation amongst all the androgen metabolites inferred a predominant hepatic route to 5 alpha-androstane-3 alpha,17 beta-diol formation, the same analogy was not applicable to the hirsute subjects. Excellent agreement was found for the predicted vs actual excretion of 5 alpha-androstane-3 alpha,17 beta-diol in normal women, based on a regression model involving the six other steroids as independent variables. When the same model was used for estimation of 5 alpha-androstane-3 alpha,17 beta-diol levels in thirteen hirsute subjects, misclassified as "normal", 50% gave values which were considerably less than actually measured. It is suggested that this discrepancy, with respect to these hirsute subjects is a reflection of extrahepatic production of 5 alpha-androstane-3 alpha,17 beta-diol due to increased 5 alpha-reductase activity.