The objective was to determine the incidence and prognosis of ventilator-associated pneumonia (VAP) in intensive care units (ICUs) in Melbourne (29-bed ICU), Australia and Aarhus and Aalborg (22-bed unit and 8-bed ICU, respectively), Denmark and to characterize participating ICUs with regard to prevalence of nosocomial type bacterial pathogens, antibiotic resistance and antibiotic consumption. In this prospective cohort study 25 patients in Melbourne and 32 patients in Aarhus + Aalborg had a first episode of VAP. The incidence of VAP per 1000 ventilator d was 6.2 in Melbourne and 9.5 in Aarhus + Aalborg. Case fatality during hospital admission was 28% and 59%, respectively (unadjusted odds ratio (OR) 0.3, 95% confidence interval (CI) 0.1-0.8). OR adjusted for age and APACHE II score was 0.2 (95% CI 0.1-1.0). Nosocomial type pathogens including methicillin resistant Staphylococcus aureus were more prevalent in Melbourne, and antibiotic consumption per VAP patient was 35% higher in Melbourne than in Aarhus + Aalborg. To judge from the present data, there seems to be a complicated interrelationship between prognosis on the 1 hand and antibiotic consumption and resistance on the other. A more favourable prognosis was found in Melbourne, where levels of antibiotic consumption and antimicrobial resistance were higher than in Aarhus + Aalborg.