OBJECTIVE: To characterize risk factors for Stenotrophomonas maltophilia bloodstream infection in oncology patients. DESIGN: A 3:1 case-control study. SETTING: Stem Cell Transplant and Leukemic Center at Barnes-Jewish Hospital (St. Louis), a 1,442-bed, tertiary-care teaching hospital with a 26-bed transplantation ward. METHOD: From June 1999 to April 2001,13 patients with S. maltophilia bacteremia were compared with 39 control-patients who were on the transplantation unit on the same day as the case-patients' positive blood cultures. Information collected included patient demographics, medical history, history of transplantation, transplantation type, graft versus host disease; neutropenia, antibiotic use, chemotherapy, mucositis, diarrhea, the presence of central venous catheter(s), cultures, and concomitant infections. RESULTS: Significant risk factors for S. maltophilia bacteremia included severe mucositis (7 [53.8%] of 13 vs 8 [20.5%] of 39; P = .034), diarrhea (7 [53.8%] of 13 vs 8 [20%] of 39; P = .034), and the use of metronidazole (9 [69.2%] of 13 vs 8 [20.5%] of 39; P = .002). In addition, the number of antibiotics used (median, 9 vs 5; P < .001), duration of mucositis (median, 29 vs 15 days; P = .032), and length of hospital stay (median, 34 vs 22 days; P = .017) were significantly different between case- and control-patients. Nine S. maltophilia isolates. tested by pulsed-field gel electrophoresis were found to be distinctly different. CONCLUSION: Interventions to ameliorate the severity of mucositis; reduce antibiotic pressure, prevent diarrhea, and promote meticulous central venous catheter care may help prevent S. maltophilia bloodstream infection in oncology patients. The role of gastrointestinal tract colonization as a potential source of S. maltophilia bacteremia in oncology patients deserves further investigation (Infect Control Hosp Epidemiol 2003;24:269-274).