Butyrivibrio fibrisolvens strains D1 and A38 produced little lactate, but strain 49 converted as much as 75% of its glucose to lactate. Strain 49 had tenfold more lactate dehydrogenase activity than strains D1 or A38, this activity was stimulated by fructose 1,6-bisphosphate, and had a pH optimum of 6.25. A role for fructose 1,6-bisphosphate or pH regulation of lactate production in strain 49 was, however, contradicted by the observations that very low concentrations (< 0.2 mM) of fructose 1,6-bisphosphate gave maximal activity, and continuous cultures did not produce additional lactate when the pH was decreased. The lactate production of strain 49 was clearly inhibited by the presence of acetate in the growth medium. When strain 49 was supplemented with as little as 5 mM acetate, lactate production decreased dramatically, and most of the glucose was converted to butyrate. Strain 49 did not possess butyrate kinase activity, but it had a butyryl-CoA/acetate CoA transferase that converted butyryl-CoA directly to butyrate, using acetate as an acceptor. The transferase had a low affinity for acetate (K(m) of 5 mM), and this characteristic explained the acetate stimulation of growth and butyrate formation. Strains D1 and A38 had butyrate kinase but not butyryl-CoA/acetate CoA transferase, and it appeared that this difference could explain the lack of acetate stimulation and lactate production. Based on these results, it is unlikely that B. fibrisolvens would ever contribute significantly to the pool of ruminal lactate. Since relatives of strain 49 (strains Nor37, PI-7, VV1, and OB156, based on 16S rRNA sequence analysis) all had the same method of butyrate production, it appeared that butyryl- CoA/acetate CoA transferase might be a phylogenetic characteristic. We obtained a culture of strain B835 (NCDO 2398) that produced large amounts of lactate and had butyryl-CoA/acetate CoA transferase activity, but this strain had previously been grouped with strains A38 and D1 based on 16S rRNA sequence analysis. Our strain B835 had a 16S rRNA sequence unique from the one currently deposited in GenBank, and had high sequence similarity with strains 49 and Nor37 rather than with strains A38 or D1.