INTRODUCTION: The clinical and economic burden of ventilator-associated pneumonia (VAP) is uncontested. We conducted the present study to determine whether low nurse-to-patient ratio increases the risk for VAP and whether this effect is similar for early-onset and late-onset VAP. METHODS: This prospective, observational, single-centre cohort study was conducted in the medical intensive care unit (ICU) of the University of Geneva Hospitals. All patients who were at risk for ICU-acquired infection admitted from January 1999 to December 2002 were followed from admission to discharge. Collected variables included patient characteristics, admission diagnosis, Acute Physiology and Chronic Health Evaluation II score, co-morbidities, exposure to invasive devices, daily number of patients and nurses on duty, nurse training level and all-site ICU-acquired infections. VAP was diagnosed using standard definitions. RESULTS: Among 2,470 patients followed during their ICU stay, 262 VAP episodes were diagnosed in 209/936 patients (22.3%) who underwent mechanical ventilation. Median duration of mechanical ventilation was 3 days (interquartile range 2 to 6 days) among patients without VAP and 11 days (6 to 19 days) among patients with VAP. Late-onset VAP accounted for 61% of all episodes. The VAP rate was 37.6 episodes per 1,000 days at risk (95% confidence interval 33.2 to 42.4). The median daily nurse-to-patient ratio over the study period was 1.9 (interquartile range 1.8 to 2.2). By multivariate Cox regression analysis, we found that a high nurse-to-patient ratio was associated with a decreased risk for late-onset VAP (hazard ratio 0.42, 95% confidence interval 0.18 to 0.99), but there was no association with early-onset VAP. CONCLUSION: Lower nurse-to-patient ratio is associated with increased risk for late-onset VAP.