A multicentre prospective cohort study was performed in 17 intensive care units (ICUs) in tertiary care hospitals in Hubei Province, China. Ventilator-associated pneumonia (VAP) was defined according to modified criteria from the published literature. Among 4155 ventilated patients, the crude incidence and incidence rate of VAP were 20.9% and 28.9 cases per 1000 ventilator-days. Multivariate analysis using logistic regression revealed risk factors including male sex [risk ratio (RR): 1.5; P < 0.001], coma (RR: 2.1; P < 0.001), chronic obstructive pulmonary disease (RR: 1.4; P < 0.001), infections at other sites (RR: 1.6; P 0.001), serious disease predating the onset of VAP (RR: 1.6; P < 0.001) and interventions including antacid treatment (RR: 1.4; P < 0.001), antimicrobial treatment (RR: 5.1; P < 0.001), bronchoscopy (RR: 1.5; P = 0.041) and tracheostomy (RR: 1.4; P = 0.014). The most frequently isolated causative pathogens were Pseudomonas aeruginosa and Acinetobacter baumannii. Of all Staphylococcus aureus isolates, 45.7% were meticillin resistant. Rates, risk factors and causal pathogens of VAP in ICUs in Hubei differ from those reported from developed countries. These data show the need for more effective infection control interventions in Hubei, China. (C) 2011 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved.