Ampicillin-resistant Enterococcus faecium (ARE) and vancomycin-resistant E. faecium (VRE) are important nosocomial pathogens. We quantified effects of probiotics and antibiotics on intestinal acquisition of ARE colonization in patients hospitalized in two non-intensive care unit (non-ICU) wards with high ARE prevalence. In a prospective cohort study with crossover design, all patients with a length of stay of >48 h were offered a multispecies probiotic product twice daily until discharge (4.5 months, intervention period) or not (4.5 months, control period). Perianal ARE carriage was determined <48 h after admission, twice weekly, and <48 h before discharge. The first isolates were genotyped by multiple-locus variable-number tandem repeat analysis (MLVA). Risk factors for acquisition were determined by Cox proportional hazards modeling, with special emphasis on ecological postantibiotic effects and delays between actual acquisition and culture positivity. Of 530 patients included, 94 (18%) were ARE colonized on admission. Of the remaining 436 noncolonized patients, 92 acquired ARE colonization: 28 (25%) of 110 probiotic users and 64 (20%) of 326 control patients (chi(2) test, P = 0.325). In all, 661 ARE strains were isolated from 186 patients, of which 186 were genotyped. In both wards, two MLVA types (MTs; MT1 and MT159) were responsible for >80% of acquisitions. Both MTs were genetically different from the probiotic E. faecium strain. Antibiotics to which ARE is resistant (hazard ratio [HR], 7.73 [95% confidence interval (CI), 4.52 to 13.22]), an ecological postantibiotic effect (HR, 7.11 [95% CI, 3.10 to 16.30]), and age (HR, 1.01 [95% CI, 0.99 to 1.02]) were associated with ARE acquisition. The HR of probiotics was 1.43 (95% CI, 0.88 to 2.34). In a setting with high selective antibiotic pressure, probiotics failed to prevent acquisition of multiresistant enterococci.