Thirteen normal subjects were investigated with a broad frequency-band (up to 4.5 Hz) rotatory test. The vestibular stimulation consisted of either a pseudo-random or a sinusoidal frequency sweep oscillation. Eye movements were recorded by EOG. Head movements were recorded either by a potentiometer attached to the chairs axis of rotation or by an angular rate sensor attached to a biteboard . Gain and phase values between eye and head movements were calculated. Measurements were performed in light and darkness during the execution of various fixation tasks. In the higher frequency range there was a in gain and phase between potentiometer and rate sensor measurements due to movement of the head relative to the headholder . Visual suppression, in agreement with earlier studies, was more effective during a sinusoidal stimulation than during a pseudo-random stimulation at low frequencies, using the more precise rate sensor measurement technique. Different instructions in the dark could alter the gain values at lower frequencies (up to 2 Hz) during sinusoidal stimulation. During pseudo-random stimulation no such differences could be elicited. Under all test conditions with the rate sensor, the gain values approached unity at about 3 Hz.