I. OBJECTIVE. The present study was undertaken in order to determine the onset of monaural deafness, especially whether it is congenital or acquired, making an investigation into the sense of sound direction in monaural hearing impairment and monaural deafness and comparing them between hearing-impaired patients and normal hearers. II. SUBJECTS. This study was carried out on 26 patients with monaural hearing impairment, 22 patients with monaural deafness and 10 normal hearers. III. METHODS. A circle with a radius of 1.3m was drawn around a fixed patient's position in a sound proof room, and the circle was divided into 16 directions at an equal angle of 22.5 degrees. A blindfolded hearer was instructed to listen to a speaker for white noise of 60dB (A) for one second and verbally answer in which direction he heard the noise. The normal hearers were tested in 4-, 8-, and 16-directions, and the hearing-impaired patients were tested in only 8-directions. IV. RESULTS. 1. Normal Hearers The rate of correct answers decreased with increasing directions of sound. The incorrect answers in 4-direction testing were only confusion between forward and backward directions, and similar incorrect answers were made in 8- and 16-direction testing. All other incorrect answers were errors of less than 45 degrees. 2. Monaurally Hearing-impaired Patients The rate of correct answers on the whole was low. There was such a relationship between the rate of correct answers and the mean hearing level of patients that the total rate of correct answers decreased with increasing hearing impairment. This correlation was statistically significant, and there was a still more significant correlation between the degree of hearing impairment and the rate of correct answers as to the right and left directions. Incorrect answers were errors of 90 degrees or less on the healthy side, while errors were made for all directions on the affected side. 3. Monaurally Deaf Patients The rate of correct answers was by far lower on the affected side. Whereas many errors were within 45 degrees on the healthy side, errors were made for all directions on the affected side. Judging from the onset of hearing loss, the monaurally deaf patients were divided into a group of 8 patients who obviously had sudden acquired deafness and a group of 9 patients who were presumed to have congenital monaural deafness. The mean rate of correct answers of the former group was superior to the latter group's, particularly on the effected side.