This study tested the potential value of routine microbiological cultures before VAP onset in predicting the causative pathogen of VAP and in anticipating the correct antibiotic choice. Authors cultured blood, respiratory secretions, catheter-tips, urine, and others substances, and found that there was no relationship between the colonization of tissues other than from the lungs and the etiology of VAP. High-quality distal bronchial samples obtained by BAL and protected-specimen brush, if cultured in the previous 72 h, had positive predictive values of 25% and 28%, respectively. Only distal respiratory secretions, when repeated at intervals between 48 and 72 h, might have a role in predicting the causative bacteria of VAP. Special attention is required for potentially drug-resistant bacteria like methicillin- resistant Staphylococcus aureus, P aeruginosa, and Acinetobacter baumannii with positive predictive values of 62%, 52%, and 24%, respectively. Previous colonization with any of these pathogens implies a high risk for subsequent pneumonia, allowing for an adequate empiric antibiotic selection.