We studied the effect of several calcium channel blockers (omega-Conotoxin-GVIA, 1 and 3microM; omega-Agatoxin-IVA, 100nM; Nitrendipine, 1 and 10microM) on evoked transmitter release at singly and dually innervated endplates of the levator auris longus muscle from three- to six-day-old rats. In dually innervated fibers, a second endplate potential may appear after the first one when we increase the stimulation intensity. The lowest and highest endplate potential amplitudes are designated "small endplate potential" and "large endplate potential", respectively. The percentage of doubly innervated junctions remains almost constant throughout the age range examined. Nevertheless, the percentage of junctions innervated by three or more terminal axons drops, whereas the singly innervated junctions increase. Therefore, between postnatal days 3 and 6, roughly half the neuromuscular junctions may experience the final process of axonal elimination. The synaptic efficacy of the large endplate potential in dual junctions, measured as the mean amplitude of the synaptic potential and mean quantal content, was the same as in the junctions that had become recently mono-innervated in the same postnatal period. In singly innervated fibers, the endplate potential size was strongly reduced by both the P/Q-type voltage-dependent calcium channel blocker omega-Agatoxin-IVA (79.17+/-4.02%; P < 0.05) and the N-type voltage-dependent calcium channel blocker omega-Conotoxin-GVIA (56.31+/-7.80%; P < 0.05), whereas endplate potential amplitude was not significantly changed by the L-type voltage-dependent calcium channel blocker Nitrendipine. In dually innervated fibers, the P/Q-type voltage-dependent calcium channel blocker omega-Agatoxin-IVA and L-type voltage-dependent calcium channel blocker Nitrendipine increased the size of the small endplate potential (161.29+/-47.87% and 109.32+/-11.03%, respectively; P < 0.05 in both cases) and reduced the large endplate potential (74.42+/-15.32% and 70.91+/-10.04%, respectively; P < 0.05 in both cases). The N-type voltage-dependent calcium channel blocker omega-Conotoxin-GVIA significantly increased the small endplate potential in the first few minutes after toxin application (at 10min: 90.23+/-17.38%; P < 0.05). This increase was not maintained, while the large endplate potential was strongly inhibited (69.25+/-7.5%; P < 0.05). In conclusion, in the dually innervated endplates of the newborn rat, presynaptic calcium channel types can have different roles in transmitter release from each of the two inputs, which suggests that nerve terminal voltage-dependent calcium channels are involved in neonatal synaptic maturation.