BACKGROUND: To identify specific risk factors for Pseudomonas aeruginosa and evaluate the relationship between the mortality rate and P. aeruginosa bacteraemia in bloodstream infections, we compared the clinical features and outcomes of patients with P. aeruginosa bacteremia with the patients with Klebsiella pneumoniae or Enterobacter bacteremia. MATERIALS AND METHODS: A total of 190 patients with P. aeruginosa bacteremia were identified from January 1998 to December 2002 and included in this retrospective analysis. During the same period, 377 patients with K. pneumoniae bacteremia and 183 patients with Enterobacter bacteremia were identified and compared with those with P. aeruginosa bacteremia. RESULTS: Factors associated with P. aeruginosa bacteremia in the multivariate analysis included pneumonia, soft tissue infection, nosocomial acquisition, neutropenia, and prior invasive procedure (All P<0.05). The 30-day mortality rate was 37.9% (72/190) in patients with P. aeruginosa bacteremia, 24.1% (91/377) in those with K. pneumoniae, and 25.7% (47/183) in those with Enterobacter bacteremia (P<0.001). However, in the analysis including patients who had received appropriate initial antimicrobial therapy (n=552), the mortality rate of P. aeruginosa bacteremia was not significantly higher than that of non-pseudomonas bacteremia (28.6% [18/63] vs. 22.5% [110/489]; P=0.282). Inappropriate initial antimicrobial therapy was found to be one of the significant independent predictors of mortality. P. aeruginosa bacteremia as a risk factor for mortality did not reach statistical significance (OR, 1.30; 95% CI, 0.73-2.32; P=0.371), after adjusting for underlying illness and adequacy of antimicrobial therapy. CONCLUSION: An initial empirical antimicrobial coverage of P. aeruginosa should be seriously considered in patients with pneumonia, soft tissue infection, neutropenia, and prior invasive procedure, when gram-negative sepsis was suspected in nosocomial infection.