The output of 11 established ileostomies was compared with ileal flow measured by intestinal perfusion in five normal volunteers when fasting and during the ileal passage of test meals containing different proportions of medium chain triglyceride and long chain triglyceride. Oroileal transit of the meal was the same in the two groups, but ileostomy output was less than ileal flow of normal persons both fasting (16.3 +/- 10.9 vs 62.4 +/- 24.7 ml/h, p less than 0.001) and after the long chain triglyceride rich meal (35.4 27.0 vs 96.1 +/- 20.2 ml/h, p less than 0.001). After ingestion of the medium chain triglyceride rich meal, ideal flow failed to increase in normal subjects but in ileostomates the changes in flow after medium chain triglyceride and long chain triglyceride rich meals were not significantly different. The fasting ileostomy effluent composition differed from that of normal fasting ileal content in having a higher concentration of potassium (8.0 +/- 2.9 vs 4.7 +/- 0.6 mmol/1, p less than 0.04) and a higher osmolality (353 +/- 63 vs 287 +/- 5 mosm/kg, p less than 0.05). Sodium concentration tended to be lower in ileostomy effluent, but in contrast to previous reports, ileostomy effluent was of consistently alkaline pH (7.2 +/- 0.3). These concentrations were not significantly altered by either type of meal. The long chain triglyceride rich meal increased the ileal flow of bile acids in both normal subjects and ileostomates, whereas the medium chain triglyceride rich meal increased bile acid flow in ileostomates but not in normal subjects, possibly reflecting a different amount of the bile acids in the ileum of the ileostomate. In the adapted ileostomate, the low volume and high potassium concentration of fasting effluent suggest that sodium and water absorption are continuously stimulated by chronic salt depletion.