An open-label randomised clinical trial was designed to compare the efficacy and tolerance of levofloxacin and ciprofloxacin plus phenethicillin for the prevention of bacterial infections in patients with high-risk neutropenia, and to monitor the emergence of antimicrobial resistance. Adult patients (n = 242) scheduled to receive intensive treatment for haematological malignancies were assigned randomly to receive oral prophylaxis with either levofloxacin 500 mg once-daily (n = 122), or ciprofloxacin 500 mg twice-daily plus phenethicillin 250 mg four-times-daily (n = 120). The primary endpoint was failure of prophylaxis, defined as the first occurrence of either the need to change the prophylactic regimen or the initiation of intravenous broad-spectrum antibiotics. This endpoint was observed in 89 (73.0%) of 122 levofloxacin recipients and in 85 (70.8%) of 120 ciprofloxacin plus phenethicillin recipients (RR 1.03, 95% CI 0.88-1.21, p 0.71). No differences were noted between the two groups with respect to secondary outcome measures, including time to endpoint, occurrence of fever, type and number of microbiologically documented infections, and administration of intravenous antibiotics. A questionnaire revealed that levofloxacin was tolerated significantly better than ciprofloxacin plus phenethicillin. Surveillance cultures indicated the emergence of viridans group (VG) streptococci resistant to levofloxacin in 17 (14%) of 122 levofloxacin recipients; in these cases, the prophylactic regimen was adjusted. No bacteraemia with VG streptococci occurred. It was concluded that levofloxacin and ciprofloxacin plus phenethicillin are equally effective in the prevention of bacterial infections in neutropenic patients, but that levofloxacin is tolerated better. Emergence of levofloxacin-resistant VG streptococci is of concern, but appears to be a manageable problem.