We now have the ability to accurately diagnose ventilator-associated pneumonia (VAP), which should lead to re-examination of nondiagnostic aspects of VAP Previous data regarding the epidemiology, treatment, and prevention of VAP were derived almost exclusively from studies in which the diagnoses were based on clinical criteria and cultures of endotracheal aspirates. These criteria are inaccurate in as many as 60 percent of cases compared to autopsy or quantitative cultures of bronchoscopic specimens. Many recent studies of VAP have focused on the accuracy of bronchoscopic and other methods of obtaining uncontaminated specimens from the lower respiratory tract for diagnosis. In anticipation of the use of these more accurate techniques to validate conclusions derived from studies based on the clinical diagnosis of VAP, the appropriateness of the design and conclusions of these published studies have been reviewed. In this communication, we discuss the areas in which sufficient data exist to reach consensus regarding methods for future studies in the epidemiology, treatment, and prevention of VAP and also attempt to delineate areas for future research endeavors.