A case-control study was conducted in a university hospital to determine the risk factors for nosocomial infection with multidrug-resistant Pseudomonas aeruginosa (MDR-PA) among all hospitalized patients and among those with a nosocomial infection due to P. aeruginosa. Eighty patients infected with MDR-PA, 75 infected with a non-MDR phenotype and 240 random controls were included in the 12-month study. Among all hospitalized patients, age, severity index, having a bedridden condition, transfer from other units, nasogastric feeding, urinary catheterization and exposure to beta-lactams (OR=2.5) or fluoroquinolones (OR=4.1) in the seven days before infection were linked to nosocomial infection due to MDR-PA. Among patients infected by P. aeruginosa, exposure to fluoroquinolones (OR=4.7) or surgery (OR=0.5) were linked to the isolation of MDR-PA. This study showed that, in addition to urinary catheterization, nasogastric feeding is an important risk factor in MDR-PA infection. Indeed, an imbalance in gut flora, modifications to the mucous membranes due to the use of nasogastric feeding and the selection pressures exerted by antibiotics were implicated in the occurrence of this infection.