Hypertransaminasemia is a frequent side effect during chenodeoxycholic administration for gallstone dissolution. Evidence suggests that this effect is not mediated by lithocholic acid, the intestinal metabolite of chenodeoxycholic acid, but that toxicity is due to the chenodeoxycholic acid itself. In vitro cytotoxicity of bile salts is positively proportional to their detergent effect, which is, on the other hand, related to their hydrophobic-hydrophilic balance. We hypothesize that in vivo also liver injury can occur when the liver is perfused by an high proportion of strongly detergent bile salts. The more detergent bile salts are unconjugated or glycine conjugated, while the lesser are taurine conjugated and sulfated. Within each class the following order of decreasing detergent power can be indicated: lithocholic greater than deoxycholic greater than chenodeoxycholic greater than cholic greater than ursodeoxycholic acid. Besides chronic exogenous administration of chenodeoxycholic or deoxycholic acids, conditions in which the liver is perfused by an high mass of highly detergent bile salts are those characterized by an enhanced intestinal biodegradation of bile salts. These conditions, which are common features of some chronic inflammatory bowel diseases, are frequently associated with liver damage. On the other hand, a normally detergent bile salt pool can become hepatotoxic for liver cells which have already been injured. In this respect, as already reported for increased sulfation, the increased proportion of taurine conjugates and the reduced formation of deoxycholic acid in liver cirrhosis can be regarded as protective mechanisms. Liver toxicity induced by bile salts' detergent action can be prevented by favouring tauroconjugation or reducing the intestinal degradation of bile salts or by administering poorly detergent bile salts.