The quantitative 14C-deoxyglucose (DG) autoradiographic technique has been used to study changes in cerebral metabolism during forelimb movements induced by graded stimulation of motor cortex. Experiments were directed at studying basic physiologic and anatomic aspects of the metabolic changes. Single shocks caused movement without metabolic change, whereas low-frequency trains caused seizures. Repetitive high-frequency train stimuli of short duration (500 Hz for 20 msec) caused jerk movements coupled with DG uptake in pathways. With stimulation of the forelimb motor zone at frequencies of 15-30/min there was prominent activation of cortical columns and strips in ipsilateral SI, SII, and MII, and contralateral MI and SI. Higher frequencies (120/min) were required to cause significant changes in DG in subcortical circuits. The most prominent changes occurred within a longitudinal corridor in dorsal thalamus and a ventral corridor in second-order sites in basal ganglia. Metabolic activation also occurred in contralateral cerebellum, the cuneate nucleus, and dorsal horn of the cervical spinal cord. Changes in these latter two sites were largely eliminated by removing feedback sensory activity. Stimulation of the forelimb sensory zone activated different sites in caudatoputamen and thalamus but similar zones in midbrain and cerebellum. The magnitude of the metabolic response in distant sites depended on the frequency of cortical stimulation. Different frequency-response relationships in different sites seemed to reflect the nature of the cortical input as well as differential effects of anesthesia. The pattern of the metabolic response was studied by comparing sites of activation with sites of the anatomic projections from motor and sensory cortical zones. 3H- and 14C-labeled amino acids were used to map the site and relative strength of pathways. Results revealed good correlation between the site of anatomic projection and the site of DG uptake but no consistent relationship between the relative strength of a projection and the magnitude of metabolic change within its field. Changes in glucose utilization with metabolic mapping experiments depend on the nature, strength, and frequency of stimulation; the site and nature of anatomic projection; the effects of anesthesia; and the strength of sensory feedback associated with the induced behavior.