A case study of a severe cholestatic syndrome induced by danazol, with ultrastructural description of liver morphology, is reported. Cholestasis appeared after 3 months treatment with danazol (300 mg/daily) and completely resolved 2 months after withdrawal. In spite of the severe increase in serum bilirubin and total serum bile salts, transaminases were only slightly elevated and GGT and alkaline phosphatases were almost normal. Light microscopy shows a pattern of predominantly centrolobular cholestasis without necrosis, with minimal inflammatory infiltrate and with no sign of bile ductule involvement. At the ultrastructural level very dilated bile canaliculus predominate with stunted or loss of microvilli and dense bile material in the lumen. Nonspecific alterations were seen in hepatocyte intracellular organelles. It is suggested that danazol may cause a rare but severe hepatocanalicular cholestasis, differing from the "bland" cholestasis frequently described during therapy with other anabolic steroids.