1. Input from group I afferents of ankle and toe extensors, other muscles, skin nerves and descending tracts to interneurones of Rexed's laminae V-VI in the cat spinal cord was analysed using intracellular recording from these interneurones. Adequate stimuli (muscle stretches) were used to activate selectively group Ia muscle spindle afferents of triceps surae and plantaris while other fibre systems were excited electrically. 2. Ia and Ib afferents of ankle and toe extensors were found to co-excite, co-inhibit or exert opposite synaptic actions in 41, 33, and 50% of the analysed interneurones, respectively. Taking into account both excitatory and inhibitory input from these two groups of afferents, 64% of the interneurones appeared to be used in common in reflex pathways from muscle spindles and tendon organs of ankle and toe extensors. 3. Selective input from Ib afferents of triceps surae and plantaris (excitation and/or inhibition) was found in 36% of the interneurones; there was evidence for a similarly selective input from Ia afferents. 4. A great majority (over 90%) of the interneurones excited by group I afferents were also inhibited by group I afferents, from either the same or other muscles. 5. Both monosynaptic and disynaptic e.p.s.p.s from Ia and/or Ib afferents from other muscles and from fibres in the ipsilateral funiculi were found in a great proportion of the same interneurones, together with disynaptic e.p.s.p.s from low threshold cutaneous afferents. 6. Intracellular staining with horseradish peroxidase revealed four different patterns of axonal projections of the analysed interneurones: (i) projections to motor nuclei and the intermediate region, (ii and III) projections only to the intermediate region, locally or combined with projections to different rostro-caudal levels, and (iv) projections to the opposite side of the spinal cord. 7. A large proportion of interneurones projecting to motor nuclei displayed input from both Ia and Ib afferents although such an input was a feature of interneurones with other projections as well. No systematic differences in the input from group I afferents were found for interneurones with different axonal projections. In contrast disynaptic e.p.s.p.s of cutaneous origin and monosynaptic e.p.s.p.s upon stimulation of ipsilateral spinal tracts appeared predominantly in interneurones projecting to motor nuclei.