1. Whole-cell patch-clamp recording has been used to study tetanus-induced synaptic potentiation of dual-component excitatory postsynaptic currents (EPSCs) in the CA1 region of rat hippocampal slices, following blockade of GABAA and GABAB receptor-mediated synaptic inhibition. 2. At a holding potential of -60 mV, the initial slope of the EPSC (between 10 and 60% of maximum amplitude) provided an accurate measurement of the AMPA receptor-mediated component, and the amplitude of the EPSC at a latency of 100 ms provided the best estimate of the size of the NMDA receptor-mediated component. 3. Neurons were voltage clamped for at least 45 min prior to delivery of a tetanus (test intensity, 100 Hz, 1 s). Measurements at 10 and 30 min following the tetanus were used as indications of short-term potentiation (STP) and long-term potentiation (LTP), respectively. One set of neurons were voltage clamped at -60 mV throughout. These neurons could be subdivided into two populations on the basis of whether or not there was LTP (n = 9), or only STP (n = 6), of the AMPA receptor-mediated component. A second set of neurons were voltage clamped at -60 mV for 30 min and then at -50 mV for 15 min before, during and for 30 min following tetanization. In these experiments there was STP but not LTP (n = 8). 4. In all neurons (n = 23), the time course of the potentiation of the NMDA receptor-mediated component paralleled that of the AMPA receptor-mediated component. In addition, potentiation of the NMDA and AMPA receptor-mediated components were of a similar magnitude. 5. These data demonstrate that it is possible to induce LTP by high frequency stimulation after 45 min of whole-cell recording. Under these conditions, there is a parallel potentiation of the AMPA and NMDA receptor-mediated components of dual-component EPSCs. This constitutes the first evidence, from studies of dual-component synaptic responses, which is consistent with a presynaptic locus of expression of tetanus-induced STP and LTP in the hippocampus.