The etiologic diagnosis of ventilator-associated pneumonia (VAP) should be considered as a microbiological emergency due to its impact on morbidity and mortality. Sampling of the lower respiratory tract (LRT) must be performed before starting or modifying antimicrobial therapy. Surveillance cultures in patients without criteria of VAP are not recommended. There is no evidence of any superiority of bronchoscopic over non-bronchoscopic sampling procedures, but quantitative bacterial cultures are essential to allow colonization to be differentiated from true infection of the LRT. Under these conditions, negative cultures practically rule out bacterial infection or, at least, identify patients who will not benefit from antibiotic therapy or who will require a very short course of treatment. Given that identification and antimicrobial susceptibility testing of microorganisms usually takes up to 3 or 4 days, rapid procedures that provide the clinician with useful information are essential. Rapid information, even if partial or less than perfect, is clearly better for the patient than a perfect but delayed report. Gram stain of LRT secretions is an immediate procedure that can guide management and it has a reasonable correlation with culture results. At present, new antibiogram procedures, performed on direct clinical samples, allow presumptive identification and information on susceptibility to commonly used antibiotics in less than 24 hours after sampling. The impact of using this procedure in clinical practice is currently under research.