1. Thalamic neurons that receive synaptic input from the globus pallidus or the cerebellar nuclei were identified in awake monkeys trained to perform an arm-reaching task. The location of electrophysiologically identified cerebellar-receiving (CR) and pallidal-receiving (PR) neurons was used to identify a total of 264 thalamic neurons in cerebellar (CB) or pallidal (GP) regions of the thalamus. 2. Stimulation in the brachium conjunctivum or white matter adjacent to the cerebellar nuclei excited 85 neurons in the thalamus at short latencies. These CR neurons were located in the oral portion of the ventral posterolateral nucleus (VPLo), in caudal portions of the ventral lateral nucleus (VLc), and in area X. 3. Stimulation in the internal globus pallidus (GPi) inhibited 10 thalamic neurons at short latency. These PR neurons were located in rostral portions of VLc, in the oral part of the ventral lateral nucleus (VLo), and in the parvicellular part of the ventral anterior nucleus (VApc). 4. There was no clear single somatotopic organization of neurons in CB and GP regions of the thalamus, as defined by "free-form" responses to passive manipulation and observation of eye movements. There was, in fact, a tendency for two representations, each, of the head/eye/mouth cells and cells with modifications of activity in response to manipulation of the arm. 5. During the hold period before illumination of a visual target, the mean firing rates and variability of discharge of arm-related CR and PR neurons did not differ significantly. This was also true for the total sample of arm-related neurons in the CB versus GP regions. 6. The activity of many neurons in both the CB and GP regions began to change before the reaching movement and, for some, before the earliest recorded changes in electromyographic (EMG) activity. The initial change was an increase in discharge for greater than 75% of the cells studied in both the CB and GP regions. 7. During the reaching task, there also was no significant difference in the time of the initial change in discharge of neurons in the CB versus GP regions of the thalamus. 8. These data are consistent with the hypothesis that the initial task-related change in discharge of PR thalamic neurons is dominated by input from the cerebral cortex and that pallidal input modulates later phases of their movement-related changes in activity.