Experiments in humans and in monkeys have indicated that load perturbations, occurring during voluntary movements and postural activity, may be automatically compensated for. Overall muscle stiffness opposing load changes is determined by the visco-elastic properties of the muscle, by segmental reflex actions and finally by long-loop reflexes. Under certain circumstances, for instance when the subject or the experimental monkey is "prepared" to counteract perturbations which are unpredictable in time, the long-loop "reflexes" appear to be responsible for most of the corrective muscle tension. Experiments in anaesthetized monkeys revealed that signals from stretch afferents reach neurons of the motor cortex, possibly via a relay in the cortical area 3a. The latencies of these responses to well controlled muscle stretches were in the same range as motor cortical cell discharges recorded in alert monkeys subjected to load perturbations. Furthermore, these responses of cells in the motor cortex also had the appropriate timing to indicate a causal relationship with the long-latency electromyographic responses to load changes referred to above. These experimental results therefore strongly support the hypothesis, first proposed by Phillips (1969), of a transcortical servo-loop adjusting motor cortical output according to the load conditions in which movements are performed. The major advantage of transcortical regulations as opposed to segmental regulations, seems to be a powerful gain control acting at the cortical level; it was repeatedly shown that the long-loop reflexes are strongly modifiable and under voluntary control. It is suggested that an adaptive gain control at the cortical level is a prerequisite to preserve the complex capabilities of the motor cortex as the chief "executive" for skilled, preprogrammed movements. A loss of this adaptive gain control may be, at least partly, the cause of motor disorders such as rigidity in Parkinsonian patients, as reported by Tatton and Lee (1975). It is suggested that further investigations of the control of transcortical reflexes may aid in the understanding of the pathophysiology of motor disabilities.