To determine the usefulness of samples obtained by bronchoscopy using a protected specimen brush and evaluated by quantitative culture techniques in establishing the diagnosis of nosocomial pneumonia in patients requiring mechanical ventilation, we prospectively studied 147 ventilated patients suspected of having nosocomial pneumonia because of the presence of a new pulmonary infiltrate and purulent tracheal secretions. Positive cultures of protected brush specimens (greater than 10(3) cfu/ml) were found in only 45 patients (31%). Subsequent follow-up confirmed the diagnosis of pneumonia in 34 of 45 patients, and in only 4 of 45 patients was a positive culture firmly established to be a false positive result. No patient with less than 10(3) cfu/ml was subsequently shown to have had pneumonia, and the diagnosis was definitely excluded in 72 of 102 patients by the absence of pneumonia at autopsy or recovery without antibiotic therapy. In contrast, when 16 clinical variables were evaluated by stepwise logistic regression analysis, no combination could be identified that was useful in distinguishing patients with bacterial pneumonia. Furthermore, when the actual costs of evaluation and therapy of our patients were compared with the projected costs entailed in treating all patients suspected of having pneumonia with antibiotics, evaluation using the protected specimen brush and quantitative cultures was less expensive after only 6 days of treatment. These results suggest that the appearance of pulmonary infiltrates and purulent tracheal secretions does not result from bacterial pneumonia in a majority of patients.(ABSTRACT TRUNCATED AT 250 WORDS)