The frequency of coughing was determined in patients with various respiratory diseases between the hours of 11:00 P.M. and 7:00 A.M. Most patients had pulmonary tuberculosis, pneumonia, or chronic obstructive respiratory disease. Coughs were counted for 3 successive nights shortly after admission to the hospital. Those with tuberculosis were studied before starting drug treatment; for other diseases, appropriate treatment had usually been insitituted before cough counting was begun. Analysis of variance showed significant variation in cough count between patients in each disease category, but variation from night to night did not exceed random variation. Median cough counts for patients with pulmonary tuberculosis, pneumonia, and chronic obstructive respiratory disease were 23, 120, and 110, respectively. In those with pulmonary tuberculosis, cough frequency varied with roentgeno graphic extent of disease. There was no significant difference in cough count between males and females, or between smokers and non-smokers, for any of the disease categories studied. Correspondence between cough frequency and awareness of cough was less than might be expected.