Optokinetic nystagmus was recorded and measured in 101 subjects comprising six diagnostic categories: 1) normal, screened for otologic disease, 2) chronic unilateral labyrinthectomy, 3) unilateral Menière's disease, 4) neurologically confirmed focal brainstem lesion, 5) brainstem-cerebellar syndrome, and 6) focal unilateral supratentorial lesion. For the OKN test, each subject looked at a translucent screen onto which a field of parallel black and white bars was back-projected. The array of bars could be projected vertically or horizontally to allow for study of nystagmus beating right and left or up and down. The speed of movement of the bars varied over a range from 20 to 140 degrees/sec of visual angle, in each axis for both directions. An analysis of the slow phase velocity of OKN indicated that patients with brainstem disease produced significantly lower eye speeds than did normal subjects or patients with chronic peripheral vestibular disease. The latter groups could not be distinguished. The responses of patients with cortical lesions fell midway between these two extremes and were significantly different from those of the brainstem group. Directional preponderance of nystagmus proved to be significantly related to the side of lesion for both the labyrinthine and cortical groups. However, the absolute value of the difference in slow phase velocity for nystagmus beating toward or away from the side of lesion was no greater than the difference between right and left-beating nystagmus in normal subjects. While the results provide statistical confirmation for the findings of earlier investigations, it is noted that for purposes of clinical diagnosis, the test is of value only in the context of the otoneurological test battery. Distribution of results for individuals in the various groups overlap considerably. The designation of a numerical cutoff for differential diagnosis leads to error rates far in excess of what may be confidently attributed to chance.