Four cannulated ewes were used in a factorial (2 x 3) design to study the effect of two sources of roughage (alfalfa hay vs ammonium-treated straw, UTS) and three levels of ground barley grain supplementation (F:C ratio: 100:0; 66:33; 33:66. respectively) on rumen digestibility, fermentation pattern and microbial N yield. Diets were restricted to maintenance level (1.6 Mcal ME/d). Rumen (0.40 vs 0.34; P=0.07) and total tract DM digestibility were higher (0.77 vs 0.68; P<0.001) when ewes were fed with alfalfa diets and digestibility increased linearly with concentrate supply (P<0.001). Alfalfa diets also promoted higher rumen ammonia (31 vs 20 mg N-NH3/100ml; P=0.003) and VFA concentrations (144 vs 82 mmol/l; P<0.001). Among levels of barley supply only the highest (F:C. 33:66) depressed significantly pH and acetate proportion (P=0.008). Protozoa counts increased significantly (P<0.05) with moderated barley supplementation (3.2·105 vs 8.2·105 vs 7.6·105 cells/ml). The highest and the lowest duodenal N flow corresponded to un-supplemented alfalfa and un-supplemented straw diets as a result of a significant (P<0.001) interaction (forage x supplementation). Moreover, rumen microbial yield was maximal when animals received non-supplemented alfalfa (9.0, 12.1 and 8.0 g/d using 15N, purine bases (PB) and DNA sequences, respectively). Using specific DNA sequences it was possible to distinguish between bacterial and protozoal N flow in duodenal digesta. Protozoal N in duodenal digesta ranged between 2.4 and 8.3 percent of total microbial flow and peaked at the highest level of grain supply (F:C 33:66). However, the possibility for DNA degradation by gastric digestion could not be ruled out. Barley supplementation promoted a linear decrease in microbial protein synthesis efficiency (g/kg OM digested in the rumen) using PB and DNA as microbial markers (37 vs 27 vs 23; P<0.05).