Cerebellar ataxic gait was compared with normals by a simple quantitative method. Strain gauges attached to the heel, the first and the fifth metatarsal areas of a shoe, were used and the gait was telemetrically recorded. A natural and one per second click sound-synchronized gait was analyzed. A contact sequence of the three areas of the sole, the stride time, the stance-phase time and the foot contact pattern were compared between patients and normals. All of these items disclosed different results in the two groups.