Two brothers with benign recurrent intrahepatic cholestasis were studied over a period of 6 years. During this period, 11 episodes of cholestasis were observed, with a mean duration of 2.6 months (range: 2 weeks to 6 months). Once, both brothers developed cholestasis simultaneously. There was a prevalence for episodes of cholestasis in wintertime. The postprandial rise in serum sulfated glycolithocholic acid was increased in the patients, and the bile acid pool was enriched with secondary bile acids. In periods prior to cholestasis, the urinary 3 alpha OH-bile acid concentration was often elevated (greater than 50 mumoles per liter) without a clear correlation with the clinical prodromata. However, it could not be used as a predictor of cholestasis. In contrast, the postprandial rise in serum 3 alpha OH-bile acids was always grossly elevated in periods just before cholestasis. An increase both in fecal bile acid excretion as well as secondary bile acids in the bile acid pool indicated an increased spillover of bile acids into the large bowel. Cholestyramine administered directly after the first signs of cholestasis appeared to shorten an episode of cholestasis. On the other hand, withdrawal of cholestyramine in a cholestasis-free period may have resulted in an episode of cholestasis. Neither taurine supplementation for 3 and 7 weeks nor calcium phosphate, which binds sulfated bile acids in vitro, for 3 weeks could prevent an episode of cholestasis, although the latter normalized the bile acid pool composition. There is a rationale for a fat-restricted diet and cholestyramine therapy only as maintenance treatment.(ABSTRACT TRUNCATED AT 250 WORDS)