Four sheep (live weight 39 to 42 kg) fitted with rumen cannulas were offered ammonia-treated barley straw alone (control) or supplemented with unmolassed sugar-beet pulp or rolled barley at 0·20 or 0·40 of the total diet on a fresh weight basis (SBP20, SBP40, B20 and B40, respectively) using a 4 × 5 randomized block design. Total diet intakes were 710, 873, 1054, 843 and 1021 g dry matter per day for control, SBP20, SBP40, B20 and B40 respectively. The intake of straw was depressed significantly with SBP40 and B40 (P < 0·05). Rumen pH and ammonia concentration decreased, and total volatile fatty acid concentrations increased (P < 0·05) with increasing level of supplementation for both supplements. Rumen liquid outflow rate was affected little by the supplementation although this measurement for B40 was significantly higher than that for the other treatments. Microbial protein supply, estimated from urinary excretion of purine derivatives, increased significantly with the increasing amounts of supplements (4·5, 6·7, 8·5, 7·1 and 9·0 (s.e.d. 0·73) g N per day for control, SBP20, SBP40, B20 and B40 respectively) (P < 0·05). The calculated efficiency of microbial protein production was 12·8, 14·2, 14·3, 15·7 and 14·5 (s.e.d. 1·45) g N per kg digestible organic matter intake (DOMI) for the five treatments. Although this value was lowest with the control, differences between treatments were not significant (P > 0·05). It seems that microbial protein yield per unit DOMI for a diet based on ammonia-treated straw could not be improved substantially by supplementation with sugar-beet pulp or barley.