Readiness potentials on voluntary hand movements were recorded from the scalp (C3: left central, C4: right central), premotor cortex, subcortical white matter and VL nucleus of the thalamus. The subjects were 6 healthy right-handed men and 23 patients with involuntary movement disorders. The subjects lay on a bed in a dark room where they performed quick, repetitive voluntary contractions with the left or right fist. The contractions were self-paced at a frequency of about one every 6 sec. Linked mastoid electrodes served as a reference throughout the experiment. The EEG was amplified with Nihonkohden RDU-5 DC amplifiers. Clenching of the fist triggered the pulse-generator which produced an immediate pulse and a one second delayed pulse. An EMG was also recorded. These signals were recorded on magnetic tape. Analysis of the data was acoomplished by playing the tape back. The EEG and EMG activity were summated by Nihonkohden ATAC 501-20 computor. In most subjects, readiness potentials were obtained before the voluntary movements. Readiness potentials were measured as the amplitude of premovement potential (N) and the interval between the beginning of the potential and the initiation of motor action (T). 1. Readiness potentials with negative shift were recorded on the scalp (C4, C3). In 6 right-handed healthy men, the means of T and N were 0.8 sec. and 7.0 μV, respectively, in C4, and 0.8 sec., 8.6 μV in C3 on right-hand movements. Conversely, the means of T and N on left hand movements were 1.0 sec., 9.2 μV in C4, and 1.0 sec., 8.4 μV in C3, respectively. The amplitude (N) in C3 on right-hand movements was significantly higher than in C4 (p<0.05 t-test). 2. Readiness potentials on the scalp were also recorded in 16 patients who had involuntary movement disorders, such as Parkinsonism, torsion dystonia or intention tremors. The means of T and N were 1.3 sec., 7.7 μV in C4 and 1.3 sec., 8.2 μV in C3 on right-hand movements. Conversely, the means of T and N on left-hand movements were 1.2 sec., 8.2 μV in C4 and 1.2 sec., 6.3 μV in C3. The mean T value was longer than the control group (p<0.01, t-test). 3. In Parkinsonian patients, the mean T value of readiness potentials on the central region contralateral to the hand movements was 1.1 sec. in 6 patients without akinesia (stage I, II). Conversely, that in 4 patients with akinesia (stage III) was 1.4 sec., which was longer than the control group (p<0.05 t-test). The length of T rather than the rigidity seemed to correlated with akinesia in patients with Parkinsonism. 4. Simultaneous recordings from the premotor cortex, subcortical white matter (2cm below the cortex) and VL nucleus of the thalamus were done during stereotactic surgery in the nonanesthetic state. Patterns of readiness potentials recorded from the premotor cortex were similar to those recorded from the scalp, but those recorded from the VL nucleus and the white matter were reversed in polarity. According to simultaneous recordings from the cortex and VL nucleus, readiness potential began approximately 0.2 sec. earlier at the cortex than the VL nucleus (p<0.01 t-test). This result is some evidence that the readiness potential initiates from the cortex of the motor area contralateral to the moving hand. 5. Readiness potentials were recorded 2-4 weeks after stereotactic VL thalamotomy. The mean T and N values of readiness potentials were almost the same as those before the surgery. The readiness potential was also recorded in a patient with thalamic syndrome who had a vascular lesion in the right thalamus. The potential showed a normal pattern from the intact side of the scalp, but on the lesion side, no potential could be obtained. These results suggest that the thalamus plays an important role in the origin of the readiness potential of the motor cortex.