We have compared state-dependent transmission through oligosynaptic (minimally disynaptic) reflex pathways from low-threshold cutaneous and muscle afferents to some flexor and extensor lumbosacral motoneurons during fictive locomotion and scratching in decerebrate unanesthetized cats. As reported in earlier work, oligosynaptic cutaneous excitatory postsynaptic potentials (EPSPs) in flexor digitorum longus (FDL) and inhibitory postsynaptic potentials (IPSPs) in extensor digitorum (EDL) longus motoneurons were enhanced markedly during the early flexion phase of fictive locomotion. We show in this paper that, in contrast, these cutaneous reflex pathways were depressed markedly during all phases of fictive scratching. On the other hand, disynaptic EPSPs produced by homonymous and synergist group I muscle afferents in flexor (tibialis anterior and EDL) motoneurons were present and strongly modulated during both fictive locomotion and scratching. During both actions, these disynaptic group I EPSPs appeared or exhibited the largest amplitude when the motoneuron membrane potential was most depolarized and the parent motor pool was active. There was an interesting exception to the simple pattern of coincident group I EPSP enhancement and motoneuron depolarization. During locomotion, disynaptic group I EPSPs in both FDL and flexor hallucis longus (FHL) motoneurons cells were facilitated during the extension phase, although FDL motoneurons were relatively hyperpolarized whereas FHL cells were depolarized. The reverse situation was found during fictive scratching; group I EPSPs were facilitated in both FDL and FHL cells during the flexion phase when FDL motoneurons were depolarized and FHL cells were relatively hyperpolarized. These observations suggest that the disynaptic EPSPs in these two motor nuclei are produced by common interneurons. Reciprocal disynaptic inhibitory pathways from group Ia muscle afferents to antagonist motoneurons were also active and subject to phase-dependent modulation during both fictive locomotion and scratching. In all but one cell tested, reciprocal disynaptic group Ia IPSPs were largest during those phases in which the motoneuron membrane potential was relatively hyperpolarized and the parent motor pool was inactive. Oligosynaptic PSPs in motoneurons produced by stimulation of the mesencephalic locomotor region (MLR) were modulated strongly during fictive locomotion but were suppressed powerfully throughout fictive scratching. Large cord dorsum potentials generated by MLR stimuli also were suppressed markedly during fictive scratching. These results allow certain inferences about the organization of interneurons in the pathways examined. They also suggest that the central pattern generators that produce fictive locomotion and scratching are organized differently.