Notes: Summary The activity in α and γ efferent axon populations and in group I and group II afferent fibre populations innervating a flexor muscle, the sartorius medialis, was observed during spontaneous locomotor movements in the thalamic cat. Multi-unit discharges of each kind of fibre were obtained by electronic sorting of the action potentials from the overall activity of a thin, intact branch of the sartorius medialis nerve. The following results were obtained: (1) The γ-motoneurones have a phasic behaviour characterized by a single discharge period during the hip flexion (swing phase of the step-cycle). (2) The γ-motoneurones are co-activated with the homonymous α-motoneurones. (3) Between rhythmic α and γ discharges, i.e. during the hip extension (stance phase of the step cycle), both α- and γ-motoneurones were normally silent. However, in 5 out of 17 experiments, a few units of the γ population fired at very low frequency. (4) Two observations indicate that the γ-motoneurones that are co-activated with the α-motoneurones by central locomotor commands are predominantly of the static type. In actual locomotion, the rhythmic fusimotor discharges over-compensate the depressor effect on the firing rate of the group II afferents of the unloading of muscle spindles by the active shortening of the parent muscle. In fictive locomotion, when the transmission of the excitation is blocked by selective curarization in alpha skeleto-motor junctions alone, the rhythmic fusimotor discharges elicit in-phase modulations not only of the group I but also of the group II fibres. The group II afferent population consists almost entirely of fibres arising from spindle secondary endings which are located primarily on intrafusal muscle fibres whose contraction is exclusively controlled by static fusimotor motoneurones. In the two experimental circumstances, the analysis of the group I fibre discharge does not allow to decide whether dynamic γ motoneurones are firing or silent during rhythmic γ discharge. (5) The group I and group II afferent discharges during the step-cycle showed two frequency peaks, one static-fusimotor dependent while the contracting muscle shortened during the hip flexion (swing) phase, the other length-change dependent while the relaxed muscle was rapidly stretched during the first part of the hip extension (stance) phase. Then, during the second part of hip extension when the muscle was slowly stretched in the absence of fusimotor drive, the firing rate of the spindle afférents decreased to a low level. The spindle sensory endings during the extension phase showed low dynamic and static responsiveness like deefferented spindles. (6) The results obtained in sartorius medialis (flexor) muscle are discussed in comparison with the results previously obtained in gastrocnemii (extensor) muscles (Bessou et al. 1986). The consequences of the predominant activation of the static or dynamic fusimotor system in functionally different muscles are considered with respect to the proprioceptive or motor role of musclespindles during muscle contraction.