To assess the potential efficacy of evernimicin (SCH 27899) against serious enterococcal infections, we used a rat model of aortic valve endocarditis established with either a vancomycin-susceptible Enterococcus faecalis or a vancomycin-resistant Enterococcus faecium strain. Animals infected with either one of the test strains were assigned to receive no treatment (controls) or 5-day therapy with one of the following regimens: evernimicin 60-mg/kg of body weight intravenous (i.v.) bolus once daily, 60-mg/kg i.v. bolus twice daily (b.i.d.), 60 mg/kg/day i.v. by continuous infusion, or 120 mg/kg/day i.v. by continuous infusion. These regimens were compared with vancomycin at 150 mg/kg/day. In animals infected with E. faecalis, evernimicin at 120 mg/kg/day by continuous infusion significantly reduced bacterial counts in vegetations (final density, 5.75 ± 3.38 log₁₀ CFU/g) compared with controls (8.51 ± 1.11 log₁₀ CFU/g). In animals infected with 0.5 ml of an 8 × 10⁷-CFU/ml inoculum of the vancomycin-resistant E. faecium, both 60-mg/kg bolus once a day and b.i.d. dose regimens of evernimicin were very effective (viable counts, 3.45 ± 1.44 and 3.81 ± 1.98 log₁₀ CFU/g, respectively). Vancomycin was unexpectedly active against infections induced with that inoculum. In animals infected with a 10⁹-CFU/ml inoculum of the vancomycin-resistant E. faecium, the evernimicin 60-mg/kg i.v. bolus b.i.d. reduced viable counts in vegetations compared with controls (6.27 ± 1.63 versus 8.34 ± 0.91 log₁₀ CFU/g; P < 0.05), whereas vancomycin was ineffective. Although resistant colonies could be selected in vitro, we were not able to identify evernimicin-resistant clones from cardiac vegetations. An unexplained observation from these experiments was the great variability in final bacterial densities within cardiac vegetations from animals in each of the evernimicin treatment groups.