SUMMIARY1. The organization of primary afferent depolarization (PAD) produced by excitation of peripheral sensory and motor nerves was studied in the frog cord isolated with hind limb nerves.2. Dorsal root potentials from sensory fibres (DR-DRPs) were evoked on stimulation of most sensory nerves, but were largest from cutaneous, joint and flexor muscle afferents. With single shock stimulation the largest cutaneous and joint afferent fibres gave DR-DRPs, but potentials from muscle nerves resulted from activation of sensory fibres with thresholds to electrical stimulation higher than 1-21-5 times the threshold of the most excitable fibres in the nerve. This suggests that PAD from muscle afferents is probably due to excitation of extrafusal receptors.3. Dorsal root potentials produced by antidromic activation of motor fibres (VR-DRPs) were larger from extensor muscles and smaller or absent from flexor muscles. The VR-DRPs were produced by activation of the lowest threshold motor fibres.4. Three types of interactions were found between test and conditioning DRPs from the same or different nerves. With maximal responses occlusion was usually pronounced. At submaximal levels linear summation occurred.Near threshold the conditioning stimulus frequently resulted in a large facilitation of the test DRP. All three types of interactions were found with two DR-DRPs, two VR-DRPs or one DR-DRP and one VR-DRP.5. The excitability of sensory nerve terminals from most peripheral nerves was increased during the DR-DRP. The magnitude of the excitability increase varied roughly with the magnitude of the DR-DRP evoked by the conditioning stimulus. D. 0. CARPENTER AND P. RUDOMIN 6. There was a marked excitability increase of cutaneous and extensor muscle afferent terminals during the VR-DRP. Flexor muscle afferent terminals often showed no excitability changes to ventral root stimulation. In those experiments where afferent terminals from flexor muscles did show an excitability increase, the effects were smaller than those of cutaneous and extensor terminals.7. The VR-DRPs appear to reflect activity of a negative feed-back loop from extensor motoneurones on to sensory fibres from cutaneous and extensor muscles. This system may have a role in modulating the ballistic movement of the frog. DR-DRPs, on the contrary, are widespread in origin and distribution. PAD from sensory fibres may function to sharpen contrast between incoming afferent information.