We developed and solved analytically a simple and general stochastic model to distinguish the univesicular from the multivesicular mode of glutamate release. The model solution gives analytical mathematical expressions for average values of quantities that can be measured experimentally. The comparison of these quantities with the experimental measures allows the discrimination of the release mode and the determination of most probable values of model parameters. The model has been validated at glutamatergic CA3-CA1 synapses in the hippocampus from newborn (P1-P5 old) rats. Our results strongly support a multivesicular type of release process requiring a variable pool of immediately releasable vesicles. Moreover, computing quantities that are functions of the model parameters, the mean amplitude of the synaptic response to the release of a single vesicle (q) was estimated to be 5-10 pA, in very good agreement with experimental findings. In addition a multivesicular type of release was supported by the following experimental evidences: (i) a high variability of the amplitude of successes, with a coefficient of variation (CV) ranging from 0.12 to 0.73; (ii) an average potency ratio a₂/a₁ between the second and first response to a pair of stimuli larger than 1; (iii) changes in the potency of the synaptic response to the first stimulus when the release probability is modified by increasing or decreasing the extracellular calcium concentration. Our results indicate that at Schaffer collateral-CA1 synapses of the neonatal rat hippocampus a single action potential may induce the release of more than one vesicle from the same release site.