Due to the greater in vitro activity of ciprofloxacin than that of levofloxacin against Pseudomonas aeruginosa, the likelihood of isolating a clinical strain of quinolone-resistant (QR) P. aeruginosa might be greater after exposure to levofloxacin than ciprofloxacin. We examined the risk of isolating QR P. aeruginosa in association with prior levofloxacin or ciprofloxacin exposure. A case-case-control study was conducted. Two groups of cases, one with nosocomial QR P. aeruginosa infections and one with nosocomial quinolone-susceptible (QS) P. aeruginosa infections, were compared to a control group of hospitalized patients without P. aeruginosa infections. Bivariable and multivariable analyses were used to determine risk factors for isolation of QR P. aeruginosa and QS P. aeruginosa. One hundred seventeen QR P. aeruginosa and 255 QS P. aeruginosa cases were identified, and 739 controls were selected. Exposures to ciprofloxacin were similar among all three groups (8% for controls, 9.4% for QR P. aeruginosa cases, and 7.5% for QS P. aeruginosa cases; P >/= 0.6). Levofloxacin use was more frequent in the QR P. aeruginosa cases than in the controls (35.9% and 22.1%, respectively; odds ratio [OR] = 2.0; 95% confidence interval [CI] = 1.3 to 3.0) and less frequent in QS P. aeruginosa cases (14.1% of QS P. aeruginosa cases; OR = 0.6; 95% CI = 0.4 to 0.9). In multivariable analysis, levofloxacin, but not ciprofloxacin, was a significant risk factor for isolation of QR P. aeruginosa (OR for levofloxacin = 1.7 [95% CI = 1.0 to 2.9]; OR for ciprofloxacin = 1.2 [95% CI = 0.6 to 2.5]). Levofloxacin was associated with a reduced risk of isolation of QS P. aeruginosa (OR = 0.6; 95% CI = 0.4 to 0.9), whereas ciprofloxacin had no significant effect (OR = 1.0; 95% CI = 0.6 to 1.8). In conclusion, the use of levofloxacin, but not ciprofloxacin, was associated with isolation of QR P. aeruginosa.