Between October, 1966 and April, 1967, 312 patients with a provisional diagnosis of pneumonia and 154 control patients without respiratory disease admitted to the Philadelphia General Hospital were studied for the purpose of determining whether or not viral infections preceded the onset of bacterial pneumonia. Of the 312 patients in the former group, only 193 were considered ultimately to have had pneumonia. Of these 193 infections, 106 were thought to have been caused by pneumococcus, 12 by Kl. pneumoniae, ten by other bacteria, and one by adenovirus type 3. In 65 patients, many of whom received treatment before hospitalization, a causal agent could not be assigned. Herpes simplex virus was isolated from 5% of cases of pneumonia and 1.4% of controls. Significant rises in antibody to influenza A and toMycoplasma pneumoniaewere detected, with one exception, only in patients with pneumonia and not in controls; but the differences were not significant statistically. Five of six patients with influenza A infection had bacterial pneumonia. The results fail to demonstrate a close association of the endemic respiratory viruses under study with bacterial pneumonia and indicate the need for development of additional virologic and serologic techniques.