1. The whole-cell configuration of the patch clamp method was used to record from hippocampal neurones in cell culture. Synaptic responses were evoked by loose patch stimulation of adjacent presynaptic neurones in low-density cultures. Agonists and antagonists were applied rapidly, using an array of flow pipes each of diameter 250 microns, positioned within 100 microns of the postsynaptic neurone. 2. Bath application of 50 microM-zinc produced prolonged periods of synaptic barrage and action potential discharge. Flow pipe application of 50 microM-zinc, in glycine-free solution with 1 mM-Mg2+, produced on average a 75% reduction of IPSP amplitude, but increased the average EPSP amplitude to 171% of control. However, after block of gamma-aminobutyric acid (GABA) receptors with bicuculline, zinc had no effect on EPSP amplitude, suggesting that potentiation recorded in control solutions reflects block of polysynaptic IPSPs. 3. Consistent with the block of IPSPs postsynaptic responses to flow pipe applications of GABA were blocked by zinc, with fast-on, fast-off kinetics. The equilibrium dissociation constant (Kd) for zinc block of GABA responses, estimated from fit of a single binding site adsorption isotherm, was 11 microM and sufficient to explain the degree of reduction of IPSPs by 50 microM-zinc. Zinc antagonism of responses to GABA was essentially independent of membrane potential over the range -60 to +60 mV. 4. With bicuculline methiodide and glycine added to a magnesium-free extracellular solution, to allow the study of synaptic responses mediated by N-methyl-D-aspartic acid (NMDA) receptors, zinc reduced the amplitude of EPSPs to 50% of control, and decreased the decay time constant of the EPSP, suggesting that zinc blocks synaptic activation of NMDA receptors. 5. Under conditions where synaptic transmission was completely blocked with postsynaptic receptor antagonists (1-3 mM-kynurenic acid and 10-20 microM-bicuculline methiodide) 50 microM-zinc decreased the amplitude of the spike after-hyperpolarization (AHP), but did not produce large changes in action potential amplitude or half-width. Under these conditions 50 microM-zinc also decreased the current threshold required to trigger action potential discharge, and blocked accommodation so that repetitive firing replaced single action potential responses to prolonged current pulses.