The effect of diet on jejunostomy output of fluid, fat, sodium, potassium, calcium, magnesium, zinc, and copper was studied in five metabolically stable, home parenteral nutrition patients. Three isocaloric diets were compared; one low in fat (30% kcal) but high in complex carbohydrate (55% kcal), and two high in fat (60% kcal) but low in carbohydrate (25% kcal). The polyunsaturated/saturated fatty acid ratios of the two high fat diets were 1:4 and 1:1. Although increasing the percentage of fat in the diet increased the amount of steatorrhea, altering the polyunsaturated/saturated fatty acid ratio had no clearly beneficial effect on the amount of fat absorbed. Neither the amount of fat, nor the type of fat, had any consistent influence on jejunostomy volume. The sodium and potassium concentration of the jejunostomy fluid stayed remarkably constant and hence net monovalent cation losses reflected jejunostomy volume rather than the fatcarbohydrate content of the diet eaten. The most consistent effect of the high fat diet was a marked increase in ostomy losses of divalent cations; calcium, magnesium, zinc, and copper. Most of the time a net divalent cation secretion on the high fat diet was converted into a net absorption on the low fat, high carbohydrate diet. Altering the polyunsaturated/saturated fatty acid ratio had no consistent effect on divalent cation losses. In conclusion, the proportion of fat versus carbohydrate calories does not appear to influence ostomy volume or monovalent cation loss in extreme short bowel, end jejunostomy patients; however, a high fat intake causes a significant net secretion of divalent cations.