The (14)C-glycocholic acid test ((14)C-GCA) has been assessed in 27 patients who have had resection of the ileum and colon for Crohn's disease and in 19 patients with unoperated stable Crohn's disease. The incidence of increased breath output of (14)CO(2) and faecal output of (14)C was significantly greater in operated patients. Half the unoperated patients had normal results and, of the others, a modest increase in faecal (14)C was the usual finding. There was no correlation between the (14)C-GCA test, the Schilling test, and the extent and severity of the radiological signs in the unoperated patients. Metronidazole therapy was poorly tolerated and had little or no influence on symptoms but did decrease the excretion of (14)CO(2) in the breath when the (14)C-GCA test was repeated. Cholestyramine therapy was beneficial in the majority of resected patients with diarrhoea and an increased faecal (14)C excretion and, on repeat testing, there was a significant increase in the output of breath (14)CO(2). The beneficial effect was less marked in the unoperated patients and the breath (14)CO(2) output remained normal in those retested. Interruption of the enterohepatic circulation of bile acids did not seem to make an important contribution to the symptoms of patients with stable unoperated (and uncomplicated) Crohn's disease, even when the ileum was extensively involved.