The coupling between presynaptic Ca²⁺ channels and synaptic vesicles is of fundamental importance for the kinetics and the efficacy of synaptic transmission. We have addressed this issue at the GABAergic basket cell–granule cell (BC-GC) synapse in the dentate gyrus. In a previous study we found that presynaptic Ca²⁺ channels and Ca²⁺ sensors of exocytosis at this synapse are tightly coupled, with distances in the range of 10 - 20 nm (Bucurenciu et al., 2008, Neuron 57:536). Here we investigated the number of Ca²⁺ channels controlling the fusion of synaptic vesicles at the BC-GC synapse. Electrophysiological BC-GC paired recordings in slices from 18- to 20-day old rats were combined with simultaneous confocal Ca²⁺ imaging. 200 mM Oregon Green BAPTA 1 and 100 mM Alexa Fluor 594 were added to the presynaptic pipette solution and putative BC terminals were visualized after 20 – 60 min. Boutons were identified as varicosities with diameter > 2 times that of the adjacent axon. The amplitude of Ca²⁺ concentration ([Ca²⁺]) transients evoked by a single action potential was <20% of that evoked by 25 action potentials, indicating that the dye quantitatively reported presynaptic [Ca²⁺] changes after single action potentials. 250 nM w-agatoxin IVa largely blocked both presynaptic [Ca²⁺] transients and evoked inhibitory postsynaptic currents (IPSCs), confirming that transmitter release at the BC-GC synapse was exclusively mediated by P/Q-type Ca²⁺ channels (Hefft and Jonas, 2005, Nature Neurosci. 10:1319). We then manipulated Ca²⁺ entry into the presynaptic terminals by reducing the extracellular Ca²⁺ concentration (which presumably changes the Ca²⁺ influx through single channels without changing channel number) or by blocking Ca²⁺ channels with w-agatoxin IVa (which presumably alters the number of available Ca²⁺ channels). When extracellular [Ca²⁺] was reduced, the relationship between transmitter release and peak amplitude of the [Ca²⁺] transient was highly nonlinear. Fitting the data with a power function (I = a [Ca²⁺]^m) revealed an exponent m of 3.21 0.18 (n = 4 pairs). In contrast, when presynaptic Ca²⁺ inflow was reduced by w-agatoxin IVa, the relationship between transmitter release and peak amplitude of the [Ca²⁺] transient was more linear, with an exponent m of 1.32 0.07 (n = 5 pairs; P < 0.05; Mann-Whitney test). Thus, transmitter release at the BC-GC synapse is initiated by a relatively small number of Ca²⁺ channels tightly coupled to the exocytotic Ca²⁺ sensors. This may contribute to the high speed and temporal precision of inhibition at BC output synapses.