To establish the incidence, etiology, risk factors, and outcomes associated with ventilator-associated pneumonia using an invasive sampling technique to avoid contamination. Eligible patients were intubated neonates treated with mechanical ventilation who followed the criteria of the Centers for Disease Control and Prevention/National Nosocomial Infection Surveillance. Bronchoalveolar lavage samples were collected using a blind-protected catheter to avoid contamination of upper respiratory microorganisms. Isolation of >10(3) colony-forming unit/mL was required for diagnosis. In 198 neonates intubated for >48 hrs, a total of 18 episodes of ventilator-associated pneumonia in 16 infants representing a prevalence of 8.1 were diagnosed. The pooled mean ventilator-associated pneumonia rate was 10.9/1,000 ventilator days. The mean age at diagnosis of ventilator-associated pneumonia was 29 ± 15 days after a mean of 21 ± 16 days of mechanical ventilation. Gram-negative bacteria were the most commonly isolated pathogens and Pseudomonas aeruginosa was the most frequent causative agent. Hospital length of stay was significantly longer for ventilator-associated pneumonia patients; however, no significant differences in mortality were found. Univariate analysis comparing patients with and without ventilator-associated pneumonia showed that days of mechanical ventilation, days of oxygen, number of reintubations, number of transfusions, bloodstream infection, and enteral feeding were all significantly associated with ventilator-associated pneumonia. However, in multivariate analysis the unique independent risk factor was days of mechanical ventilation (odds ratio 1.12, confidence interval 95% 1.07-1.17). Ventilator-associated pneumonia is a frequent nosocomial infection in newborns. Only duration of mechanical ventilation has been identified as an independent risk factor for ventilator-associated pneumonia. The use of a blind invasive sampling technique seems to diminish sample contamination.