Patients with Crigler-Najjar syndrome and Gunn rats cannot form bilirubin glucuronides owing to a lack of bilirubin UDP-glucuronosyltransferase activity. Because increased serum and tissue bilirubin levels remain constant, an alternative excretory route has to substitute for this deficiency. Gunn rats excrete in bile only 2-13% of the bilirubins eliminated in Wistar rats. In contrast, the biliary excretion rate of urobilinogen in Gunn and Wistar rats is comparable. The sum of bilirubins and urobilinogen excreted in the bile of Gunn rats amounts to 10-30% of pigments excreted in Wistar rats. Despite this low biliary excretion, the intestinal content and fecal excretion of bile pigments in Gunn and Wistar rats were similar. These data support an extrabiliary entrance of unconjugated bilirubin into the intestine. Additional proof for this was found in that the intestinal lumen of Gunn rats still contains a high amount of bilirubins and urobilinogen after 3 d of external biliary drainage. A similar procedure in Wistar rats resulted in the complete disappearance of bile pigments from the intestine. The direct transmural transport of bilirubin from blood to all parts of the intestinal lumen was demonstrated by injecting 14C-bilirubin i.v. into Gunn rats with isolated parts of small and large intestine. In Crigler-Najjar and Gilbert's syndrome patients, the biliary excretion of bile pigments has previously been shown to be strongly reduced. Their stools, however, contained approximately the same amount of bile pigments as in normal subjects. Although only traces of unconjugated bilirubin were detected in the stool of normal persons (4 +/- 3% of total bile pigments), higher amounts were found in patients with Crigler-Najjar disease (20 +/- 12&). These results suggest a direct intestinal permeation of unconjugated bilirubin in severe unconjugated hyperbilirubinemia both in man and rats.