High sodium channel densities at nodes of Ranvier are essential for action potential (AP) propagation in myelinated axons, safeguarding long-range signalling. Although voltage-gated calcium channels (VGCCs) are present in unmyelinated and premyelinated axons and in axon initial segments, and are thought to mediate a calcium-activated potassium current (IKCa:1,2,3), they have rarely been observed at nodes of Ranvier and little is known about their function. We have investigated the presence and functional role of activity-dependent calcium transients at the nodes of Ranvier of cerebellar Purkinje cells (PCs). While PCs express axonal VGCCs early in postnatal development, it is unknown whether they are retained post-myelination, outside of synaptic specialisations. Using two-photon Ca2+ imaging in cerebellar slices of P18-P43 mice we observed spatially-restricted (~5µm), activity-dependent Ca2+ influx at branch points of PC axons, which are typically nodes of Ranvier. Baseline nodal [Ca2+] was dependent on spontaneous action potential (AP) frequency, and Ca2+ transient integrals correlated with the number of APs showing that nodal [Ca2+] is directly related to neuronal activity. Rise times of AP-evoked nodal Ca2+ transients were comparable to those at the axon initial segment and soma, but nodal decay times were > two-fold faster. Ca2+ transients were blocked by removal of extracellular Ca2+ and by nickel at high (400 µM, to 14 ± 6% of control, n=4) but not low (50 µM 91±9% of control n=3) concentrations, implying involvement of R-type calcium channels. Other, more selective VGCC antagonists, however, were ineffective. Simultaneous patch clamp recordings of somatic APs and axonal capacitive currents arising from the propagating APs showed that local application of calcium-free solution to branch points robustly blocked action potential propagation (100% failure of axonal spikes, n=3), indicating that nodal calcium entry is essention for AP propagation in PC axons. Local application of antagonists for VGCCs (Ni2+/Cd2+) had variable effects reducing axonal AP propagation velocity and in some cases leading to propagation failure (4/13 cells). We hypothesised that these effects on propagation resulted from reduced recruitment of a IKCa. Although antagonists of SK and BK channels had no effect, local application of TRAM-34 (500 nM), which has been shown to selectively block the intermediate calcium-activated potassium channel, KCa3.1 (4), reduced the amplitude of axonal capacitive currents, indicating a slowing of the rise time of axonal APs and reduced sodium channel availability. KCa3.1 channels have recently been shown to modulate EPSP summation in cerebellar Purkinje cells (5), and these data show that they are also involved in securing cerebellar action potential output.