1. Intracellular recording techniques were used to investigate the mechanisms underlying the activity-dependent lability of inhibitory synaptic potentials indirectly evoked in CA3 pyramidal neurons by stimulation of the mossy fiber afferent pathway in organotypic slice cultures of hippocampus. 2. Repetitive stimulation (3-10 Hz, 30-60 s) was found to reduce the amplitude of the inhibitory postsynaptic potential (IPSP) and occasionally lead to repetitive, epileptiform discharge. 3. Under single-electrode voltage-clamp, the current underlying the inhibitory postsynaptic potential (IPSC) was found to have the same reversal potential (EIPSC) as the response to iontophoretically applied gamma-aminobutyric acid (EGABA), and both were blocked by bicuculline. Reducing the extracellular Cl- concentration from 153 to 89 mM shifted EGABA in the depolarizing direction by 9 mV from -64.7 to -55.6 mV, an amount close to that predicted by the Nernst equation. We therefore presume that the IPSC is mediated by GABA and that the reversal potentials of both are equal to ECl-. 4. Under single-electrode voltage-clamp, repetitive stimulation (3-10 Hz, 30-60 s) was found to cause a mean decrease in the conductance underlying the IPSC (gIPSC) of 22%. This decrease was independent of the membrane potential at which stimuli were delivered. 5. Under single-electrode voltage-clamp, repetitive stimulation (3-10 Hz, 30-60 s) was found to cause a 2-8 mV depolarizing shift in EIPSC when the membrane potential was held constant 5-15 mV depolarized from EIPSC. The mean decrease in IPSC driving force was 49%. If membrane potential was held 10-20 mV hyperpolarized from EIPSC, there was no change in driving force. 6. Currents activated by iontophoretically applied GABA were decreased in amplitude following repetitive stimulation at depolarized, but not hyperpolarized, holding potentials. 7. The decrease in IPSC driving force following repetitive stimulation at depolarized holding potentials was less after decreasing the extracellular K+ concentration from 5.8 to 1 mM. 8. We conclude that the decrease in driving force following repetitive stimulation results from an increase in the intracellular Cl- concentration, and that the activity-dependent decrease in gIPSC results from a decrease in presynaptic release rather than from postsynaptic receptor desensitization.