Diets containing vegetable tannins, predominantly hydrolysable gallotannins, at levels of 13.5, 25 and 50 g/kg were fed to growing broiler cockerels to examine their effect on enzymes in the pancreas, the intestinal lumen and the intestinal mucosa. Pancreas weight per unit live weight showed a significant (P < 0.05) increase with increasing level of dietary tannin while that of the liver remained unaffected. Trypsin (EC 3 . 4 . 2 1 .4) and a-amylase (EC 3.2.1.1) activities in the pancreas of birds fed at the highest level of tannins were more than double those from birds fed on a tannin-free control diet. In the intestinal lumen inhibition of trypsin activity increased with increasing level of dietary tannin; a-amylase activity was inhibited at intermediate tannin levels but was restored at the highest level. Dipeptidase (EC 3.4.13.11) and sucrose a-glucosidase (disaccharidase) (EC 3.2.1.48) in the intestinal mucosa were both inhibited by tannins. Growth of the birds and digestibility of nitrogen were adversely affected by the tannin-containing diets.Tannin : Pancreatic enzymes : Brush border enzymes : Cockerel It is well established that tannins are potential protein precipitants (Hagerman & Butler, 1980;Hagerman & Klucher, 1986; Makker et al. 1987) and that they reduce the digestibilities of proteins (Mohammed & Ahnied, 1987) when present in animal feeds. Elevated faecal nitrogen excretion associated with ingestion of tannin-containing feeds is ascribed largely to interactions between either tannins and dietary protein or tannins and digestive enzymes, or both. However, Mitjavilla et al. (1977) concluded that the excess faecal N is mostly a result of mucus hypersecretion.Studies in vitro (Griffiths, 1981 ;Lumen & Salamat, 1980;Horigome et al. 1988) and in vivo (Griffiths & Moseley, 1980; Horigome er al. 1988) have demonstrated the formation of tannin-enzyme complexes. However, adopting a technique in vitro more relevant to the situation in vivo, Mole & Waterman (1987) showed that trypsin retained all its activity in the presence of tannic acid when the system included a substrate protein (bovine serum albumin ; BSA) and glycocholic acid, indicating that under such conditions tannins have a preferential affinity for dietary protein over enzyme protein. In previous work Mole & Waterman (1985) had shown that the formation of complexes between tannins and BSA, used as a substrate for proteolysis, could result in either enhancement or inhibition of proteolysis, or no effect at all depending on the tannin:BSA ratio in the complex.In contrast to what had been reported with rats (Griffiths & Moseley, 1980; Horigome et ul. 1988), tannins were shown to exert no effect on protein digestion by insect herbivores (Martin et ul. 1985(Martin et ul. , 1987 in terms of changes in activities of enzymes in the lumen of the small intestine, in order to provide further insight into the mode of action of dietary tannins. A further aim was to examine the effect of dietary tannins on the activity of membrane-bound enzymes wh...