A study of the fourth sound was conducted on 100 normal subjects (ages 1-88 years) and 42 clinical cases with either aortic stenosis, systemic hypertension or coronary heart disease. This study was based on the graphic recognition of a presystolic sound when the tracing was taken with the use of one or more of 5 different high pass filters. Attention was paid to the existence of the fourth sound, its magnitude, and its vibrational frequency. In general it was accepted that a magnitude of 1/2 of the first heart sound or a frequency of 30 Hz denoted a pathologic fourth sound. However, exceptions were found among normal subjects, so that only the combination of the two criteria could be considered highly significant for a pathologic phenomenon (gallop). Patients with aortic stenosis presented an increase in magnitude of the fourth sound but incidence and vibrational frequency were similar to those of controls. Patients with hypertension had a greater incidence of fourth sounds, especially in middle age (100%); middle age patients usually had a greater magnitude while older patients had more often an increase in vibrational frequency. Patients with coronary heart disease (evidence of old infarcts) had an increase in the incidence, magnitude, and vibrational frequency in comparison with controls. These data and the cause of the fourth sound are discussed. The fourth sound has been repeatedly studied in the past, both as an auscultatory finding and a graphic phenomenon. Attempts were made for separating the normal fourth sound from that denoting a pathological phenomenon but, so far, no clear cut criteria for the differentiation have been obtained. We thought, therefore, that a new study was indicated.