To determine the effect of total parenteral nutrition (TPN) and different enteral feeding formulas on survival and liver function following hemorrhage in rats. Prospective randomized controlled study. Laboratory of a large university-affiliated medical school. Sixty-seven male Sprague-Dawley rats weighing 350 to 450 g. Jugular and gastroduodenal feeding catheters were inserted in animals 1 day before hemorrhage, and animals were started on one of six different fluid or nutritional regimens: TPN, iv saline, an enteral amino acid-based formula (AA) (Vivonex-TEN), an enteral peptide-based formula (PEP) (Reabilan-HN), an enteral intact-protein based formula (PRO) (Osmolite-HN), or enteral saline. A catheter was inserted in the tail artery and animals were hemorrhaged 5 mL/kg at baseline and 1 hr later. Animals were returned to their cages and observed for survival. Liver function was determined by measuring circulating bile acid levels at baseline and 24 hr after hemorrhage. Mortality was significantly increased in animals receiving TPN (63%) and AA (24%). Mortality was 13% in animals receiving PRO and 0% in animals receiving PEP and saline. Liver function deteriorated in all animals after hemorrhage except the PEP group. TPN and AA increased mortality in animals after hemorrhage. PEP was associated with zero mortality and protection of liver function after hemorrhage. (Crit Care Med 1990; 19:54)