1. Ryegrass (Lolium perenne cv. Cropper) silage was given to four Friesian steers (initial live weight (LW) 172 kg) alone or with a fishmeal supplement (150 g fresh weight/kg silage dry matter (DM)) in a balanced two-period change-over design. The dietary components were the same as those used in a recent experiment by Gill et al. (1987). All diets were offered hourly at 24 g DM/kg LW. 2. Fishmeal supplementation increased dietary nitrogen intake (P less than 0.01) and significantly increased the flow of total N (P less than 0.01), non-ammonia N (NAN) (P less than 0.01) and amino acids (P less than 0.05) at the duodenum. The increased supply of NAN to the duodenum was due largely (67%) to increased flow of undergraded dietary protein. 3. Microbial protein production was estimated simultaneously with 15N, diaminopimelic acid (DAPA) and a novel technique using L-[4,5-3H]leucine. Estimates varied with the marker and source of microbial isolate but mean values indicated that microbial N flow was significantly increased by fishmeal supplementation (P less than 0.05). The use of L-[4,5-3H]leucine as a microbial marker is justified and its possible advantages over other markers are discussed. 4. The efficiency of microbial protein synthesis was significantly increased from 30.8 g N/kg organic matter apparently digested in the rumen (OMADR) to 54.3 g N/kg OMADR by fishmeal supplementation (P less than 0.01). However, this indicates that relatively high efficiencies can be achieved with unsupplemented high quality silage supplied continuously. Rumen degradable N (RDN) supply was significantly increased by fishmeal supplementation (P less than 0.05) but apparent efficiency of capture of RDN by rumen microbes was not significantly increased. 5. Attempts were made to investigate the source of N utilized by the microbes on the two diets by intrarumen infusions of (15NH4)2SO4 and L-[4,5-3H]leucine but these were confounded by rumen-mixing problems. Findings obtained suggest that a lower proportion of microbial N may have been derived from rumen ammonia when the silage was supplemented with fishmeal but no differences in the extent of direct incorporation of leucine into microbial protein were observed. This could indicate an increase in microbial peptide uptake on the fishmeal-supplemented diet. However, evidence was also obtained suggesting that the improvement in microbial protein synthetic efficiency with supplementary fishmeal was also due to the provision of a more continuous supply of nitrogenous substrates for microbial growth, as a result of hourly feeding.(ABSTRACT TRUNCATED AT 400 WORDS)