Synaptic NMDARs provide voltage-dependent depolarization and a Ca²⁺ influx signal allowing downstream activation of Ca²⁺-dependent processes. In tandem with AMPARs which mediate most fast glutamatergic EPSCs, NMDARs mediate a slow EPSC in this dual component glutamatergic synapse. Synaptic NMDAR currents are large during early development but decline in magnitude as synaptic pathways mature. This has lead to the suggestion that NMDAR have no role at mature auditory brainstem synapses. We have shown that NMDAR-mediated nitric oxide production occurs at the mature (P35) calyx of Held synapse (1). Therefore, the present study characterises the changing properties of NMDAR from P9 to P35 and shows that they do contribute to synaptic signalling at the mature synapse. CBA/CaJ mice (P9-P35) were killed by decapitation in accordance with the UK Animals (Scientific Procedures) Act 1986. Brainstem slices containing the superior olivary complex were prepared as previously described (2) and patch clamp experiments were performed at 36°C on principal neurons of the medial nucleus of the trapezoid body (MNTB). Synaptic activity was achieved by midline stimulation using a bipolar electrode. We confirm the decline in NMDAR-mediated EPSC magnitude by P18, and show that this maturation process can be divided into two phases. First, there is an early decline in NMDAR current amplitude (P9: 3.7±0.8nA to P35: 0.3±0.1nA*, n=4-5) and acceleration in kinetics, so that NMDAR-mediated EPSCs decay with a dominant fast time-constant (P9: 36.3±3.3ms to P35: 16.7±4.1ms*, n=4-5) at mature synapses. Quantitative PCR showed increased NR2A mRNA expression after P10, consistent with faster kinetics and a declining contribution from NR2B subunits, since EPSCs from P18/P21 showed no block by the NR2B-prefering antagonist ifenprodil (10μM, Ctrl: 0.48±0.06nA vs ifen: 0.44±0.1nA, n=4). Second, the NMDAR-mediated EPSC exhibits decreased voltage-dependent block by [Mg²⁺]_o, consistent with incorporation of NR2C subunits into synaptic channels. In addition, sensitivity to extracellular Zn²⁺ was not detected, and the Zn²⁺ chelating agent, TPEN had no effect. We conclude that NMDARs are not eliminated from the calyx of Held synapse, but that channel expression evolves from an immature state where large conductance channels with slow kinetics are replaced by small conductance channels with fast kinetics and reduced Mg²⁺ block. These results confirm the relevance of NMDAR channels throughout development and suggest that the declining NMDAR-EPSC is a shift from ‘gross’ electrical signalling to a Ca²⁺-dependent intracellular signalling (e.g. activation of tightly coupled neuronal nitric oxide synthase). Results are reported as mean±SEM. Significance was tested using two-tailed Student’s t-test. Differences were considered statistically significant (*) at p<0.05.