1. The background activity was observed in gamma and alpha efferent fibres and in group I and II fibres innervating the muscle gastrocnemius lateralis or medialis. The reflex effects of ipsilateral and contralateral sural nerve stimulations on the muscle efferents were analysed together with their consequences upon the afferents of the same muscle. The observations were made in the decerebrated cat without opening the neural loops between the muscle and the spinal cord.2. The multi-unit discharges of each category of fibres were obtained, on line, by an original electronic device (Joffroy, 1975, 1980) that sorted the action potentials from the whole electrical activity of a small branch of gastrocnemius lateralis or medialis nerve according to the direction and velocity of propagation of the potentials.3. The small nerve may be regarded as a representative sample of different functional groups of fibres conducting faster than 12 m.sec(-1) and supplying gastrocnemius muscles.4. Some gamma efferents were always tonically firing except when a transient flaccid state developed. Usually the alpha efferents were silent, probably because the muscle was fixed close to the minimal physiological length.5. Separate and selective stimulations of Abeta, Adelta and C fibres of ipsilateral and contralateral sural nerve showed that each group could induce the excitation of gamma neurones. The reciprocal inhibition period of alpha efferents during a flexor reflex was only once accompanied by a small decrease in gamma-firing.6. The reflex increase of over-all frequency of gamma efferents resulted from an increased firing rate of tonic gamma neurones and from the recruitment of gamma neurones previously silent. When the gamma efferents in the small nerve naturally occurred in two subgroups, the slower-conducting subgroup (mainly composed of tonic gamma axons) was activated before the faster-conducting subgroup (mostly composed by gamma axons with no background discharge). Some rare exceptions were found, however.7. The selective activation of gamma efferents could be obtained with short-and low-frequency stimulation. When, with stronger stimulations, gamma-alpha co-activation was observed, the onset of the gamma-firing increase preceded by 100-600 msec that of the alpha discharge in the small nerve. Likewise, the onset of the gamma-efferent response preceded the increase of over-all electromyographic activity of the whole triceps muscle but only by 10-100 msec. The discrepancy could be due to the soleus alpha motoneurones being activated earlier than the alpha-motoneurones of gastrocnemius muscle, according to the size principle. In only one experiment, the alpha-firing onset preceded the gamma-firing increase.8. Stimulations of ipsilateral or contralateral nerve, whatever the alpha or gamma reflex patterns, always led to increased firing rates of group I and II fibres of the small nerve. The origins of the discharge of group I and II muscle afferents and the excitation mechanisms of the receptors involved are considered. Some aspects of the mechanism of the reflex control of movement are discussed in the light of these results.