The ruminococci are an important group of fibrolytic bacteria inhabiting the rumen. Seventeen strains of presumptively identified Ruminococcus were evaluated by a combination of nearly complete and partial 16S rDNA sequence that identified all strains as either Ruminococcus albus or Ruminococcus flavefaciens. All sequences fell into cluster IV of the Clostridia, while other species of ruminococci (eg. Ruminococcus obeum, Ruminococcus gnavus, Ruminococcus lactaris) fall into cluster XlVa of the Clostridia. Ruminococcus cluster IV sequences were used to design a 16S rRNA oligonucleotide probe to assess the relative abundance of target populations in a stable ruminal environment. A stable population (animals fed eight times per day) was established in sheep so that statistically robust comparisons could be made in the absence of variation due to diurnal rumen fluctuations. The steady state populations were sampled six times over a 24 d period and direct microscopic counts (DC), total culturable counts (TCC), and total cellulolytic counts (CEL) were determined. DC and culturable data (TCC and CEL) were compared with relative abundance estimates of Ruminococcus IV and Fibrobacter succinogenes. A combination of the Ruminococcus and F. succinogenes probes accounted for 4·0% of the bacterial population and cellulolytic bacteria (measured by most-probable numbers) were 5·2% of the total culturable count. These data suggest that a major portion of the Ruminococcus and Fibrobacter diversity has been cultured and is represented by available sequences. Steady state populations were measured over several days in three sheep and an estimate of variation in DC, TCC, CEL and 16S-based data were obtained. These variance estimates could be used to determine the theoretical sample sizes required to obtain statistically significant differences under different experimental conditions.