BACKGROUND We observed an increased rate of Pseudomonas aeruginosa bacteremia in our hematology unit in 2004-2007 without an identified environmental source. METHODS We conducted a matched case-control study to investigate factors associated with P aeruginosa bacteremia in patients with hematologic malignancies. RESULTS Forty-two episodes of P aeruginosa bacteremia were identified. At presentation, 26 patients (62%) had pneumonia and 9 patients (21%) were in shock. Twenty-five patients (60%) were aplastic. The clinical cure rate was 40%. Comparing the 42 cases with 84 matched controls identified the following independent risk factors for P aeruginosa bacteremia: hospitalization in the previous 3 months (odds ratio [OR], 12.84; 95% confidence interval [CI], 2.98-55.18), antibiotic therapy in the previous 3 months (OR, 5.34; 95% CI, 2.14-13.30), receipt of ceftriaxone in the previous 3 months (OR, 2.38; 95% CI, 1.08-5.27), receipt of aminoglycosides in the previous 3 months (OR, 6.65; 95% CI, 1.15-38.25) and receipt of fluoroquinolones in the previous 3 months (OR, 3.22; 95% CI, 1.48-7.00). CONCLUSIONS Local antibiotic therapy algorithms were modified to decrease prescriptions of ceftriaxone and combination therapy with aminoglycosides and fluoroquinolones in an effort to decrease the risk of P aeruginosa bacteremia.