In humans, electrical stimulations of various lower limb nerves helped reveal powerful excitation produced by group II afferents in motoneurones supplying knee and ankle muscles. The distribution of heteronymous group II spinal connections showed that group II from intrinsic foot muscles activate both ankle dorsiflexors and knee muscles, whereas group II from ankle dorsiflexors activate knee extensors and those from ankle plantar flexors, knee flexors. Since these combinations correspond to muscle synergies used during walking or required to maintain unstable upright stance, transmission of group II excitation from ankle to knee muscles has been investigated during various motor tasks. During both walking and standing, group II excitation is enhanced whereas group I excitation transmitted through the same interneurones was not changed as compared to that observed during tonic and voluntary co-contractions of the same muscles at matched levels of electromyogram (EMG) activity. More recent experiments, using cortical stimulation, reveal group II connections between muscles acting at the ankle joint (tibialis anterior to gastrocnemius medialis and reciprocally; see Pierrot-Deseilligny & Burke, 2005). It has been extensively discussed whether the pathways transmitting long latency reflexes originating from muscle stretches are transcortical (activated by group Ia afferents) or spinal (group II and/or cutaneous afferents; see Marsden et al. 1983; Matthews, 1991). The electrophysiological expertise concerning group II spinal reflexes developed in lower limb investigations has been extended to upper limb. Since the intensity of the electrical stimulation required for activating group II afferents needs to be above motor threshold (MT) and given the distribution of recurrent inhibition, which can mask hypothetical group II excitations, the study has been focused on the effects of ulnar nerve stimulation at wrist level in order to activate group II afferents from hand muscles. This stimulation has been used to condition the motor units’ (MUs) firing (studied with the post stimulus time histogram, PSTH method), extracted from surface EMG of forearm and arm muscles. It has been observed that ulnar nerve stimulation produced monosynaptic group Ia excitations in the PSTHs of flexor digitorum superficialis (FDS) MUs and non-monosynaptic (mediated by propriospinal neurones) group I excitation in the PSTHs of flexor carpi radialis (FCR) MUs (see Pierrot-Deseilligny & Burke, 2005). When the stimulus intensity was above 1.2 x MT, a late peak of increased MU firing appeared in the PSTH of (i) 21 out the 28 FDS MUs (all 6 subjects) so explored, with a mean latency of 10.6 ± 0.3 ms (mean ± S.E.M.; after the monosynaptic group Ia peak) and of (ii) 20 out the FCR 24 MUs (all 7 subjects, but one) so explored, with a mean latency of 11.2 ± 0.4 ms. It was checked that these peaks were not evoked when using weaker stimuli (0.8 x MT) and after pure cutaneous stimulation (external aspect of the fifth finger), which eliminate a possible contribution of low threshold afferents (group I and cutaneous). The threshold for evoking the late peak suggests the contribution of smaller diameter afferents, probably group II afferents. To test this hypothesis, the modifications in the latency of the monosynaptic group Ia and late and higher threshold peaks were assessed during cooling the ulnar nerve (ice pack placed against the palmar aspect of the forearm, along the nerve). In all 6 FDS MUs so studied (5 subjects), the late and higher threshold peak was significantly more delayed than the monosynaptic group Ia peak (8.0 ± 0.5 vs 2.6 ± 0.4 ms, respectively; p < 0.01, Mann-Withney U test). Moreover, the effects of ulnar nerve stimulation on FCR on-going EMG was tested after tizanidine oral intake (±₂ noradrenergic agonist, inhibiting group II afferent transmission; 150 ¼g kg^-1). Under these conditions, after drug intake (45-90 min), the amount of non-monosynaptic group I excitation was not statistically changed (39.9 ± 6.6 vs 35.0 ± 8.8% of the mean unconditioned EMG before drug intake) whereas that of the late and higher threshold facilitation was significantly decreased (11.0 ± 2.4 vs 44.6 ± 9.9% before drug taking; p < 0.05 Wilcoxon matched-pairs signed-rank test). These results suggest that group II afferents from hypothenarian muscles produce excitation in motoneurones involving finger and wrist flexors. It was also possible to evoke a similar late and high threshold excitation in all the motor nuclei so investigated (extensor digitorum, triceps and biceps brachii), except in motoneurones supplying wrist extension. In contrast, median nerve stimulation at wrist level evoked low threshold and short latency inhibition (43/67 MUs) or late and low threshold excitation (17/67 MUs), both mainly reproduced by pure cutaneous stimulation.