Vancomycin-resistant enterococci (VRE) have become important causes of nosocomial infections. This study evaluated the association between a variety of intravenous antimicrobial exposures and the isolation of VRE using two control groups: (1) a vancomycin-susceptible enterococci (VSE) group, to assess factors associated with development of VRE, and (2) a nonenterococci control group, to assess factors associated with positive cultures for enterococci without regard to vancomycin resistance. After adjusting for the effect of other antimicrobials, time at risk, and patient morbidity, compared to vancomycin-susceptible enterococci controls, exposures to imipenem (OR = 4.9, 95% CI = 1.6-14.1) and ceftazidime (OR = 2.6, 95% CI = 1.1-6.1) were significant predictors of VRE. When compared to nonenterococci controls, exposures to ampicillin (OR = 20.1, 95% CI = 1.5-263.1) and imipenem (OR = 5.1, 95% CI = 1.5-17.1) were significantly associated with VRE. Neither piperacillin nor vancomycin was associated with VRE compared to either control group. This study offers further evidence that the replacement of broad-spectrum cephalosporins by extended-spectrum penicillins, specifically piperacillin, may be effective in reducing VRE.