The in vitro activities of trimethoprim (TMP), alone and in combination with sulfamethoxazole (SMX), against 131 clinical isolates of enterococci, 126 Streptococcus faecalis isolates, and 5 Streptococcus faecium isolates were determined by a broth microdilution method with Mueller-Hinton broth that was substantially free of inhibitory substances. The geometric mean MIC of TMP for strains of S. faecalis was 0.164 micrograms/ml (range, 0.03 to 8 micrograms/ml), with a geometric mean MBC of 0.298 micrograms/ml (range, 0.063 to 8 micrograms/ml). Although all strains were resistant to the sulfonamide alone, the inhibitory and bactericidal activities of TMP against strains of S. faecalis were markedly potentiated when TMP was combined in a fixed ratio of 1:19 with SMX; the geometric mean MIC of TMP was reduced to 0.016 micrograms/ml (range, 0.002 to 0.25 micrograms/ml), with a geometric mean MBC of 0.031 micrograms/ml (range, 0.004 to 0.25 micrograms/ml). The combination had no synergistic effect against strains of S. faecium; the geometric mean MICs and MBCs of both agents were ca. 0.06 micrograms/ml. The MBC/MIC ratios for TMP and TMP-SMX were less than or equal to 16 for all 131 strains. MICs and MBCs for TMP-SMX were unchanged, and for TMP they decreased when performed in broth supplemented with 50% heat-inactivated pooled human serum. For TMP and TMP-SMX, the susceptibilities of isolates with high-level resistance to gentamicin or streptomycin were the same as those of isolates susceptible to less than or equal to 2,000 micrograms of aminoglycoside per ml. These results suggest that TMP-SMX and TMP alone could prove useful in the treatment of serious enterococcal infections, including infections by strains with high-level resistance to aminoglycosides.