The sources of an outbreak of Enterobacter cloacae bloodstream infections in a pediatric hospital were investigated, as were the risk factors for acquiring the infection: Two retrospective case-control studies were conducted. The study sample included all patients admitted to the general pediatric wards from February 5 through March 30, 2001, who had a positive blood culture for E. cloacae. Pediatric ward and pharmacy infection-control practices were reviewed, personnel and environmental cultures were obtained, and pulsed-field gel electrophoresis (PFGE) molecular typing of the bloodstream isolates was conducted. Four subjects were identified. These infants were more likely than control patients to receive i.v. ranitidine (p < 0.01). Among patients receiving i.v. ranitidine, subjects were more likely than controls to receive i.v. ranitidine prepared by a pharmacist. No environmental or personnel cultures yielded E. cloacae. Patients' E. cloacae isolates had four different PFGE patterns, suggesting environmental rather than point-source contamination. Ranitidine multidose vials were kept connected to an automatic compounding machine for up to 48 hours at room temperature after the first dose was drawn, contrary to manufacturer recommendations. Further, preparation of ranitidine infusions was not conducted in accordance with recommendations for risk level 2 sterile i.v. products. The use of contaminated ranitidine multidose vials was the most likely cause of an outbreak of E. cloacae. However, a combination of other factors such as inadequate hand-washing techniques, presence of E. cloacae in the environment, noncompliance with guidelines for the preparation of sterile infusions and medications, and a susceptible population may have contributed to the infections.