Voltage gated potassium (Kv) channels are critical to regulation of neurotransmitter release throughout the nervous system, but the roles and identity of the subtypes involved remain unclear. Here we show that Kv3 ion channels participate in regulation of transmitter release at the neuromuscular junction (nmj). Adult C57BL mice were killed humanely by CO2 exposure and the flexor digitorum brevis (FDB), transversus abdominus and lumbrical muscles dissected out and either aldehyde-fixed for light and electron microscopy or used for electrophysiological recordings. Light microscopic immunohistochemistry using antibodies raised in rabbit to specific sequences of Kv3.1-Kv3.4 subunits (Alomone Labs, 1:100-1:1000) revealed Kv3.3 and Kv3.4, but not Kv3.1b or Kv3.2, at neuromuscular junctions. The immunoreaction was colocalised with SV2 (Iowa hybridoma bank, 1;500) and was therefore likely to be within presynaptic terminals. Presynaptic localisation was confirmed by electron microscopy. To examine the roles of these Kv3 subunits intracellular recordings were made of evoked end plate potentials (EPPs) in FDB muscle fibres. Muscles were maintained in Ringers solution (composition in mM: NaCl, 120; NaH₂PO₄, 0.4; d-glucose, 5; NaHCO₃, 23.8; CaCl₂,2; MgCl₂,1 and KCl, 5) aerated with a mixture of 95% O₂/5% CO₂ and maintained at room temperature (22-25 °C). The tibial nerve was stimulated via a suction electrode at a frequency of 0.25 Hz with pulses of sufficient magnitude to produce a maximal twitch of the muscle (1 ms pulse duration; amplitude 0.5-2V). MgCl₂ (0.5-5 mM) or µ-Conotoxin GIIIB (2.5 µM; Alomone Labs) was added to the bathing medium to prevent muscle twitching during recordings and to reduce the size of the EPP to allow any enhancement to be observed. Tetraethylammonium applied in the presence of the BKCa channel blocker iberiotoxin (25-125 nM) and at low concentrations (0.05-0.5mM), which block Kv3 subunit containing channels specifically, did not affect muscle fibre resting potential but significantly increased the amplitude of EPPs (p<0.05, mean±SEM; 0.05 mM: 6.4±1.8 to 8.5±1.9mV (n=12); 0.1 mM: 5±0.9 to 7.7±1.2mV (n=6); 0.5 mM: 12.9±8.3 to 20.1±10.1mV (n=3)) Unexpectedly, blood depressing substance (BDS-I), a toxin selective for Kv3.4 subunits, had no effect at 0.05-1µM. The combined presynaptic localisation of Kv3 subunits and enhancement of EPP amplitude indicate that Kv3 channels influence neurotransmitter release from presynaptic terminals. The channel at the motor nerve terminals may be unusual as it is insensitive to the Kv3.4 subunit specific toxin BDS even though Kv3.4 subunits are present.