Ventilator-associated pneumonia is responsible for approximately half of the infections acquired in the intensive care unit and represents one of the principal reasons for the prescription of antibiotics in this setting. Invasive diagnostic methods, including bronchoalveolar lavage and/or protected specimen bronchial brushing, could improve the identification of patients with true bacterial pneumonia and facilitate decisions of whether to treat. These techniques also permit rapid optimization of the choice of antibiotics in patients with proven bacterial infection, once the results of respiratory tract cultures become available, based on the identity of the specific pathogens and their susceptibility to specific antibiotics, to avoid prolonged use of a broader spectrum of antibiotic therapy than is justified by the available information. Because unnecessary prolongation of antibiotic therapy for patients with true bacterial infection may lead to the selection of multidrug-resistant microorganisms without improving clinical outcome, efforts to reduce the duration of therapy for nosocomial infections are also warranted. An 8-day regimen can probably be standard for patients with ventilator-associated pneumonia. Possible exceptions to this recommendation include immunosuppressed patients, patients who are bacteremic or whose initial antibiotic therapy was not appropriate for the causative microorganism(s), and patients whose infection is with very difficult-to-treat microorganisms and show no improvement in clinical signs of infection.