1. Evoked synaptic potentials and currents were recorded in neonatal rat fourth deep lumbrical muscle during the period of polyneuronal innervation. Signs of inhibitory interactions between converging mononeurone terminals were detected. 2. Muscle fibres innervated by axons from the lateral plantar nerve (LPN) and from the sural nerve (SN) were studied. In unblocked preparations the muscle contracted, and electrode tips were mounted on flexible wires to prevent loss of impalements. 3. In voltage recordings from unblocked preparations, paired two-shock stimulation of one nerve revealed synaptic depression: the second response was smaller than the first. When the two stimuli were delivered to different nerves (SN and LPN), the second response was smaller than its own control. 4. In voltage clamped, unblocked preparations, a similar result was obtained. Conditioning stimulation of one nerve (SN, for example) inhibited the response to test stimulation of the other nerve (LPN). The inhibition was greater with larger conditioning responses, was maximal when the conditioning and test stimuli were approximately superimposed, and decayed with a time course of several tens of milliseconds. Several tests showed that the end-plate was well clamped: the observed inhibition could not be explained by voltage escape at the end-plate. 5. The inhibition was not constant during the tail of the test end-plate current (EPC). Instead, it declined during the EPC tail, suggesting that the mechanism of inhibition was active, though diminishing, throughout the time course of the test EPC. 6. The amount of inhibition was not noticeably affected by altering the muscle membrane potential (two cells studied). 7. Treatment with curare or alpha-bungarotoxin to block most ACh receptors reduced the inhibition. In about half of the fibres studied, no inhibition was evident: in the others, up to 50% inhibition was observed. The average inhibition for all receptor-blocked fibres was about 15%. 8. In six alpha-bungarotoxin-treated cells, multiple conditioning stimuli were delivered. In most cases, the amount of inhibition increased with increasing numbers of conditioning stimuli. 9. Several possible mechanisms of inhibition are discussed, including reduction of current through ACh channels during the test response owing to alterations in synaptic cleft ion concentrations produced by the conditioning response, presynaptic inhibition of transmitter release, and postsynaptic receptor saturation.