1. Whole-cell patch-clamp recordings were used to study paired-pulse facilitation (PPF) of the lateral perforant path input to the dentate gyrus in thin hippocampal slices. 2. Orthodromic stimulation of the lateral perforant pathway evoked a excitatory postsynaptic current (EPSC) with a latency of 3.3 ± 0.1 ms (mean ± SE) that fluctuated in amplitude. The EPSC had a rise time (10-90%) of 2.79 ± 0.06 ms (n = 35) and decayed with a single exponential time course with a time-constant of 9.14 ± 0.24 ms (n = 35). No correlation was found between the amplitude of the EPSC and the rise time or decay time-constant. The non-N-methyl-D-aspartate (NMDA) antagonist 6-cyano-7-nitro-quinoxaline-2,3-dione completely blocked the EPSC whereas the NMDA antagonist D-aminophosphonovaleric acid (APV) had modest effects. 3. When a test (T-) EPSC was preceded at an interval of 100 ms by a conditioning (C-) EPSC, a significant increase in the amplitude of the T-EPSC was seen in 38 out of 44 trials analyzed from a total of 27 granule cells. The average amount of PPF was 35.7 ± 2.1%. There was no apparent correlation between the amount of PPF and the stimulation intensity or mean amplitude of the C-EPSC. The time course of the facilitated T-EPSC was not significantly different from that of the C-EPSC. 4. No correlation was found between the amplitude of the C-EPSC and that of the T-EPSC. Estimates of quantal content (m(cv)) were determined by calculating the ratio of the squared averaged EPSC amplitude (from 48 responses) to the variance of these responses (M2/σ2) whereas quantal amplitudes (q(cv)) were estimated by calculating the ratio of the response variance to average EPSC amplitude (σ2/M). PPF was found to be associated with an average increase in m(cv) of 64.8 ± 7.2% (n = 38) whereas q(cv) was decreased by 12.1 ± 3.8%. 5. The time course of PPF was studied by varying the interval between the C- and T-pulse from 10 to 400 ms while keeping the stimulation intensity constant. Maximal facilitation of the T- EPSC was obtained with interpulse intervals ≤25 ms where the average facilitation amounted to ~70% (n = 6). The decline of facilitation was nearly exponential and was no longer evident with intervals >350 ms. 6. Blocking NMDA-receptors with APV had no effect on the amount of PPF indicating that only non-NMDA-receptors were involved. Bicuculline also was without effect on both the amount and the time course of PPF indicating that PPF is not because of a concomitant paired-pulse depression of γ- aminobuturic acid (GABA)ergic inhibition. Baclofen, a GABA(B)-receptor agonist, reduced the amplitude of both the C-EPSC and the T-EPSC by a similar amount. 7. When extracellular calcium was lowered, C- and T-EPSC amplitudes were reduced by an average of 62 and 55% (n = 3), respectively. This was followed by a decrease in the amplitude fluctuation of both EPSCs. The differential effect on the C-EPSC and T-EPSC resulted in an average increase in PPF of 24.6%. 8. The statistical analysis of EPSC variability, together with the results obtained by changing extracellular calcium, strongly suggests that PPF of the lateral perforant path input to the dentate gyrus is because of an increase in presynaptic transmitter release. Furthermore this increase in transmitter release seems principally to result in an increased activation of non-NMDA-receptors on the granule cells.