The effect of quebracho tannin (QT) on the activities of rumen microbes was studied both in vivo and in vitro. The in vivo study was carried out in four rumen environments created by feeding four fistulated sheep a basal diet of chopped grass hay, grass cubes and whole barley (control) with or without the addition of QT and the dietary modifier Browse Plus (BP). The active ingredients in BP are polyethylene glycol (PEG, more than 0·95 of the dry material), polyvinyI pyrrolidone (PVP) and calcium hydroxide. The diets offered were control, control plus 50 g QT per kg diet (CQT), control plus 50 g QT per kg and 0·1 (CQT+ 0·1) or 1.0 g BP per kg (CQT + 1·0). The experiment was a 4 X 4 Latin-square design with periods of 4 weeks. Rumen liquor samples were collected at 0, 2, and 8 h after feeding, 18 and 25 days after commencement of each period. Changes in free and bound proanthocyanidin (PA) concentrations, rumen pH, ammonia, volatile fatty acids (VFA), soluble protein concentrations, particle-bound cellulase and xylanase activities and rumen protozoal counts were used as indices of microbial activities. The same diets were incubated with rumen contents in vitro, except that the 0.1 g/kg level of BP was replaced by a fourth treatment made up of control plus 1·0 g BP per kg DM of hay (CBP). In vivo, bound PA concentrations in the rumen liquor of all the sheep given QT-containing diets were higher than those of the free PAs. The bound PA was higher (V < 0·05) in the rumen liquor of sheep given CQT in comparison with those containing BP. There were reductions in ammonia (P > 0·05), soluble protein (P < 0·05) and total VFA (P > 0·05) concentrations in the rumen liquor of sheep given CQT in comparison with those given the control and BP-containing diets. Reductions (P > 0·05) were also observed in the cellulase and xylanase activities of the rumen microbes and in protozoal counts (P < 0·05) in the presence of QT in the diet. In vitro, QT reduced (P > 0·05) the viable bacterial counts, gas production and the total VFA concentrations after incubation. In both the in vivo and in vitro experiments, BP was able to alleviate some but not all of the depression in microbial activities. The main constituent (PEG) in BP binds to some of the tannins in solution and may displace tannins from tannin-protein complexes.