Our aims were to identify (i) risk factors associated with the acquisition of multidrug-resistant (MDR, to 3 or more classes of antimicrobials) Proteus mirabilis isolates responsible for bloodstream infections (BSIs) and (ii) the impact on mortality of such infections. Risk factors for acquiring MDR P. mirabilis BSIs were investigated in a case-case-control study; those associated with mortality were assessed by comparing survivors and nonsurvivors in a cohort study. The population consisted of 99 adult inpatients with P. mirabilis BSIs identified by our laboratory over an 11-year period (1999 to 2009), 36 (33.3%) of which were caused by MDR strains, and the overall 21-day mortality rate was 30.3%. Acquisition of an MDR strain was independently associated with admission from a long-term care facility (odds ratio [OR], 9.78; 95% confidence interval [CI], 1.94 to 49.16), previous therapy with fluoroquinolones (OR, 5.52; 95% CI, 1.30 to 23.43) or oxyimino-cephalosporins (OR, 4.72; 95% CI, 1.31 to 16.99), urinary catheterization (OR, 3.89; 95% CI, 1.50 to 10.09), and previous hospitalization (OR, 2.68; 95% CI, 10.4 to 6.89). Patients with MDR P. mirabilis BSIs received inadequate initial antimicrobial therapy (IIAT, i.e., treatment with drugs to which the isolate displayed in vitro resistance) more frequently than those with non-MDR infections; they also had increased mortality and (for survivors) longer post-BSI-onset hospital stays. In multivariate regression analysis, 21-day mortality was associated with septic shock at BSI onset (OR, 12.97; 95% CI, 32.2 to 52.23), P. mirabilis isolates that were MDR (OR, 6.62; 95% CI, 16.4 to 26.68), and IIAT (OR, 9.85; 95% CI, 26.7 to 36.25), the only modifiable risk factor of the 3. These findings can potentially improve clinicians' ability to identify P. mirabilis BSIs likely to be MDR, thereby reducing the risk of IIAT--a major risk factor for mortality in these cases--and facilitating the prompt implementation of appropriate infection control measures.