Background/purpose Carbapenem-resistant Pseudomonas aeruginosa infections have been a challenge and issue in hospital settings. However, the clinical impact of P. aeruginosa blood isolates resistant only to carbapenems has never been discussed previously. Methods To assess the risk factors and clinical significance of bacteremia caused by carbapenem resistance only P. aeruginosa (CROPA), a 6-year retrospective case–control study was conducted. The CROPA strains were defined as isolates susceptible to ciprofloxacin, antipseudomonal penicillins and cephalosporins, and aminoglycosides but resistant to one antipseudomonal carbapenem (imipenem or meropenem) or both. The controls were selected among patients with bacteremia due to P. aeruginosa susceptible to all above classes of antipseudomonal antibiotics, which was defined as all-susceptible P. aeruginosa. Results Twenty-five patients had at least one blood culture positive for CROPA, and 50 controls had all-susceptible P. aeruginosa bacteremia. CROPA bacteremia had a high 30-day mortality rate (72.0%), as compared to 26.0% for the controls ( p  < 0.001). Through multivariate analysis, carbapenem exposure was the only risk factor for developing CROPA bacteremia ( p  = 0.002). A comparison between the surviving and deceased patients with CROPA bacteremia showed that nine (50%) of those who died, but none of the survivors, received carbapenems as the initial empirical therapy ( p  = 0.027). Conclusion Carbapenem exposure was associated with emergence of CROPA infections. Repeated carbapenem use in such patients might increase rates of inappropriate initial empirical treatment and mortality. Prudent carbapenem use is important to reduce the emergence of CROPA.