Two experiments of Latin square design were made, each with four Friesian bull calves fitted with re-entrant duodenal and ileal cannulas at 4-10 d of age. The calves were used to study the effect of giving milk-substitutes containing 0, 300, 500 and 700 g bacterial protein (Pruteen)/kg total protein on apparent digestibility of nitrogen fractions and amino acids and true digestibility of 3H-labelled milk protein and 35S-labelled bacterial protein in the small intestine. A third experiment of Latin square design with four intact Friesian calves was used to measure apparent digestibility of nutrients throughout the alimentary tract and retention of N, calcium and phosphorus. At the duodenum, volume of outflow, its pH, and outflow of total-N(TN), protein-N (PN) and non-protein-N (NPN) decreased with time after feeding. At the ileum, volume of outflow and TN outflow were unaffected by time after feeding but PN outflow decreased; NPN outflow at the ileum increased to a maximum 6 h after feeding and then declined. Increased inclusion of Pruteen did not affect the volume of outflow at the duodenum or ileum, but duodenal PN outflow increased. At the ileum, pH values were lower and TN, PN and NPN outflows were higher with increasing concentration of Pruteen in the diet. Apparent digestibility in the small intestine tended to decrease with greater amounts of Pruteen, but was only significant for NPN. Apparent digestibility from mouth to ileum significantly decreased for TN and PN as Pruteen inclusion increased. Amino acid concentration in duodenal outflow, with the exception of that of arginine, reflected intake. The total amount of each amino acid in ileal outflow increased and the apparent digestibility of most amino acids decreased with greater amounts of Pruteen in the diet. Apparent digestibility of nucleic acid-N from Pruteen was very high. True digestibility in the small intestine and between mouth and ileum of 3H-labelled milk protein was high and did not differ between dietary treatments. True digestibility of 35S-labelled Pruteen was low for the milk-protein diet and tended to increase linearly as more dietary Pruteen was included. Dry matter concentration in faeces and a high apparent digestibility throughout the whole alimentary tract of carbohydrates did not differ between treatments.(ABSTRACT TRUNCATED AT 400 WORDS)