Nystagmus-eliciting effect of the head shaking method, a test of vestibular function, was studied on 163 patients who visited our otologic clinic in last year with complaints of dizziness and distur bance of equilibrium.The method consists in shaking the patient's head 30 times(a) in the Frankfort horizontal plane with the head bent forward 30 degrees, (b) in the same horizontal plane at normal head position, (c) in the coronal plane, or(d) in the sagittal plane, each stopping suddenly in midline position.The nystagmus thus elicited was observed under Frenzel's spectacles. The nystagmus was observed in 86.1% of the patients when all of the above mentioned head shaking methods were applied to them, and in 83.0% when the only two methods(b) and(d) were applied.These incidences were exceedingly higher than those of the other methods of nystag- mus provocation such as tests for positional nystag. mus.If one desires much simplicity one can use (b) and(d) instead of all the four. In normal subjects a small horizontal nystagmus was elicited althouth infrequently, i.e.in about 14% of them.But, according to the author's experiences, the elicited nystagmus was considered as a form of pathologic sign if it was coarse in amplitude, vertical and oblique in form, or larger in the number of beating. The great importance of the head shaking test for the detection of latent imbalance of vestibular function as well as for the differentiation of the patients suffering from dizziness and disturbance of equilibrium was stressed.