The minimal inhibitory concentrations (MICs) of sulfamethoxazole and trimethoprim against 144 strains of obligately anaerobic bacteria were determined on Diagnostic Sensitivity Test agar (Oxoid) or in prereduced Diagnostic Sensitivity Test broth, both supplemented with sodium pyruvate (1 mg/ml), hemin (5 mug/ml), and vitamin K(1) (1 mug/ml). Fifty-eight percent of the strains were susceptible to sulfamethoxazole alone (MIC </= 16 mug/ml), only 12% were susceptible to trimethoprim alone (MIC </= 1 mug/ml), and 85% were susceptible to the combination of sulfamethoxazole plus trimethoprim (MIC </= 16 mug/ml) at a ratio of 19:1. All 45 strains of the Bacteroides fragilis group were susceptible to the combination. Synergy of the combination was often observed by a checkerboard MIC determination of 123 strains, usually most markedly when the ratio of the two components was near 1:1. However, there was also synergism at the ratio of sulfamethoxazole to trimethoprim of 16:1 in 61 (53.5%) of the 114 strains that could be evaluated for synergistic activity. When the strains were tested by the broth-disk test proposed by Wilkins and Thiel, modified by using prereduced Diagnostic Sensitivity Test broth instead of brain heart infusion broth and by using a smaller inoculum, there was over 90% correlation with the MICs. Poor results were found when the broth-disk tests were performed in brain heart infusion broth. There was very poor correlation between inhibition zone diameters by an agar diffusion method and MICs.