Ventilator-associated pneumonia (VAP) occurs in approximately 25% of patients undergoing mechanical ventilation. The impact of VAP on patient outcome has been much debated; however, Fagon and colleagues reported an attributable mortality of 27%. The optimal management of patients with suspected VAP requires confirmation of the diagnosis and identification of the responsible pathogen(s) to provide appropriate antimicrobial therapy. The diagnosis of VAP remains one of the most difficult clinical dilemmas in critically ill mechanically ventilated patients. Many conditions may either obscure or give rise to a clinical picture similar to that of pneumonia in the mechanically ventilated patient. A fever, pulmonary infiltrate, and purulent tracheobronchial secretions are almost diagnostic of pneumonia in a previously healthy nonintubated patient. However, this constellation of features, while suggestive, is not diagnostic of pneumonia in mechanically ventilated patients. Using multivariate analysis, Fagon and colleagues demonstrated that no combination of clinical, laboratory, or radiographic findings could accurately predict which patients had VAP.