Information processing in the brain is fundamentally constrained by the kinetics of rapid synaptic responses. In turn, the time course of excitatory postsynaptic currents has been shown to depend critically on neurotransmitter diffusion in the synaptic cleft. Most central synapses, however, are not accessible for direct measurements of neurotransmitter concentration because of their small size. To address this issue, we have focused on common excitatory glutamatergic synapses in the hippocampus (area CA1 in stratum radiatum) and combined two-photon microscopy, patch-clamp electrophysiology and biophysical modelling to assess the speed of glutamate diffusion in the synaptic cleft. First, we measured the effective diffusion coefficient of small, cell-impermeable soluble ‘probe’ molecules, the fluorescence indicator Alexa Fluor 350, in area CA1 of acute rat hippocampal slices using rapid two-photon excitation imaging of a diffusion point-source. We have found that perfusing the slices with a 5% solution of 40kDa dextran at 35°C reduces extracellular diffusivity of the probe by 25 ± 3% (n = 30). This reduction was indistinguishable from that in a free medium [1] indicating that interactions with the extracellular milieu play little role in the diffusion-retardation effect of large dextran macromolecules on small, rapidly diffusing molecules. This suggested that dextran should slow down extracellular diffusion of glutamate molecules to a similar degree. Second, we found that dextran application increased the amplitude of evoked (by 16 ± 8%; n = 21), miniature (by 13 ± 7%; n = 6) and minimum stimulation (by 14 ± 4%; n = 8) AMPA receptor-mediated EPSCs recorded in CA1 pyramidal cells (the average increase, 15 ± 4%). Third, we carried out in-depth simulations of glutamate release, diffusion and receptor activation in the typical environment of these synapses using two complementary modelling approaches, multi-compartmental and Monte-Carlo. Taken together, simulation results have proposed a robust non-linear relationship between the amplitude of AMPA receptor mediated responses and glutamate diffusivity in the cleft, as suggested earlier [2-4]. According to this relationship, a 15% reduction of the response amplitude by 25% retardation of diffusion corresponds to a unique value of the glutamate diffusion coefficient inside the cleft (currently estimated in the region of 0.3 μm²/ms). Subject to the ongoing tests, this value provides a fundamental constrain on interpreting rapid molecular events in the synaptic cleft.