During voluntary movement sensory feedback from the moving limb may contribute to drive the muscle activity, correct the ongoing movement, and update the central motor commands for future movements. The sensory feedback may be modulated in various ways to ensure that it is optimally adjusted to the requirements of the performed motor task. It has been demonstrated that much of this modulation is explained by changes in presynaptic inhibition of the synapses of Ia afferents on spinal motorneurones (Hultborn et al. 1987a,b; Meunier & Pierrot-Deseilligny, 1989; Nielsen & Kagamihara, 1993b). In recent experiments we have demonstrated that the amount of presynaptic inhibition changes in relation to the recent motor activity of the subject. In a first study, 19 healthy volunteers were asked to train a visuo-motor skill task involving the ankle muscles. The position of the ankle joint was measured by a goniometer and displayed as a cursor on a computer screen located in front of the subject. Subjects were instructed to make the cursor follow a series of figures as accurate as possible by performing voluntary ankle dorsi- and plantarflexion movements. A session consisted of 8 sets of 4 min of training with 2 min of rest in between the sets. Motor performance (measured as the deviation of the ankle joint position signal from the target trajectory) was significantly improved following the visuo-motor skill task but not after a control session consisting of simple voluntary dorsi- and plantarflexion movements. The slope of the H-reflex recruitment curve and the H-max/M-max ratio were depressed after repetition of the visuo-motor skill task and returned to baseline after 10 min. No changes were observed after the control session. To elucidate the mechanisms contributing to the H-reflex depression we measured the size of the long-latency depression of the soleus H-reflex evoked by peroneal nerve stimulation (D1 inhibition) and the size of the monosynaptic Ia facilitation of the soleus H-reflex evoked by femoral nerve stimulation. The D1 inhibition was increased and the femoral nerve facilitation was decreased following the visuo-motor skill task, suggesting an increase in presynaptic inhibition of Ia afferents. In a second study, 7 healthy human subjects had their ankle joint immobilized by a cast for a period of 2 weeks. Immediately following the immobilization, the size of the soleus H-reflex was significantly increased, but returned to pre-immobilization levels within a couple of days. This increase was caused in all likelihood by a decrease of presynaptic inhibition of soleus Ia afferents. In summary, these observations suggest that presynaptic inhibition of Ia afferents may be both down- and up-regulated depending on the preceding motor activity of the subject. The functional significance of these changes is unclear, but we believe that they may have some relevance for the hyperexcitability of the stretch reflex arch in patients with lesions of central motor pathways.