Control of vancomycin-resistant Enterococcus faecium (VRE) in European hospitals is hampered because of widespread asymptomatic carriage of VRE by healthy Europeans. In 2000, our hospital (The University Medical Center Utrecht, Utrecht, The Netherlands) was confronted with a large outbreak of VRE. On the basis of genotyping (by pulsed-field gel electrophoresis), epidemic and nonepidemic VRE strains were distinguished, and infection-control measures were exclusively targeted toward epidemic VRE. The outbreak was retrospectively divided into 3 periods of different infection-control measures. Compliance with use of alcohol-based hand rubs was enforced during all periods. Period I involved active surveillance, isolation of carriers, and cohorting (duration, 4 months); preemptive isolation of high-risk patients for VRE colonization was added in period II (7 months); and cohorting and preemptive isolation were abandoned in period III (18 months). When the outbreak was identified, 27 patients in 6 wards were colonized; 93% were colonized with an epidemic VRE strain. Detection rates of nonepidemic VRE were 3.5%, 3.0%, and 2.9% among 683, 810, and 977 screened patients in periods I, II, and III, respectively, comparable to a prevalence of 2% (95% confidence interval [CI], 1%-3.5%) among 600 nonhospitalized persons. The relative risks of detecting epidemic VRE in periods II and III, compared with period I, were 0.67 (95% CI, 0.41-1.10) for period II and 0.02 (95% CI, 0.002-0.6) for period III. Infection-control measures were withheld for patients colonized with nonepidemic VRE (76 [54%] of 140 patients with a test result positive for VRE). Use of alcohol-based hand rubs increased by 31%-275% in outbreak wards. Genotyping-targeted infection control, isolation of VRE carriers, enhancement of hand-hygiene compliance, and preemptive isolation successfully controlled nosocomial spread of epidemic VRE infection.