Ventilator-associated pneumonia (VAP) is a common nosocomial infection causing significant morbidity and mortality. The goal of this study was to determine the efficacy of bilateral versus unilateral bronchoalveolar lavage (BAL) for the detection of the causative bacterial agents of VAP. We retrospectively studied the quantitative bacterial cultures of 399 BAL sample pairs collected from 287 mechanically ventilated patients over a 5-year period at a U.S. tertiary-care teaching hospital. Trauma was the underlying illness in 69% of patients. No evidence of bacterial infection was found in 226 BAL pairs (56.6%). Among 173 positive BAL sample pairs, significant bacterial counts were detected exclusively in 6.4% of left-lung and 12.1% of right-lung samples. In contrast, 81.5% of positive sample pairs had significant bacterial counts in both lungs. All bacteria recovered at significant concentrations from bilateral samples would have been detected in a unilateral right-lung sample in 89% of positive sample pairs. Unilateral sampling would have failed to recover one or more significant isolates in 11% of positive pairs had only the right lung been sampled and in 16.7% had only the left lung been sampled. Our study shows that preferential sampling of the right lung improves the diagnostic efficacy of unilateral BAL for the detection of the etiologic agents of VAP. If bilateral sampling is performed, our results also indicate that pooling left- and right-lung samples for a single quantitative culture is comparable to processing samples individually.