In a patient with total colectomy and only 120 cm of the proximal small bowel remaining after resection for Crohn's disease, comparative studies were performed on the absorption of isocaloric amounts of Vivonex HN, Flexical, Codelid, Jejunal, and Precision LR. These elemental and/or complete diets were fed by a nasogastric tube at a constant rate of 260 ml/hr, 1 cal/ml, 2185-2660 ml per feeding period of 8.4-10.2 hr daily. Discharge of intestinal fluids from the jejunal stoma ranged between 2940 ml (Vivonex HN) to 4760 ml (Jejunal) per daily feeding period, resulting in a net intestinal loss of fluids and dehydration with the exception of Flexical. Jejunostomy discharge on Flexical was only 1325 ml per comparable period. The patient tolerated dietary fat relatively well absorbing 61% of 87 g of fat from Flexical and 60% of 108 g from a natural blenderized diet. The other diets used were either fat-free or with a minimum amount of fat. The second best tolerated diet was Vivonex HN, the feeding of which resulted in the highest intestinal retention of nitrogen (84% of the 16.3 g fed as opposed to 78% of 9.1 g fed in Flexical). Vivonex HN also had the highest intestinal retention of phosphorus. Intestinal fluid absorption was not related to the intestinal transit time measured by a nonabsorbable marker or to the osmolality of diets. Diets containing fat may inhibit secretion by the "enterogastrone-like" effect induced by dietary fat in the remaining bowel. High amino acid content of some of the low-fat diets may release gastrointestinal hormones which stimulate secretion, and the simple carbohydrates prevent fluid absorption. In the short-bowel syndrome with normal functions of the pancreaticobiliary system, diets based on fat and protein hydrolysate should be the nutritional therapy of choice.