1. γ-Aminobutyric acid (GABA) (10-500 μM) was applied to cultured chick cerebral neurons by pressure ejection, and the resulting currents (I(GABA)) were recorded using standard whole-cell voltage-clamp techniques. Plots of the peak I(GABA) as a function of membrane potential were nonlinear with and outwardly rectifying appearance. 2. I(GABA) decayed during a prolonged application of GABA. This decay was associated with a decline in the conductance of the cell, suggesting that the decline in I(GABA) was principally due to receptor desensitization. 3. After 5-7 days in culture, whole-cell recordings revealed the presence of spontaneous synaptic currents. These currents were presumed to be GABA-gated inhibitory postsynaptic currents (IPSCs) because they reversed at the Cl- equilibrium potential (E(Cl-)), were blocked by picrotoxin (25 μM), and were prolonged by pentobarbital (50 μM). 4. Synaptic currents were analyzed by fitting exponential functions to their decay. In normal recording saline, 68% of the decays analyzed could be adequately described by a single exponential function. Two exponentials were necessary to describe the decay of the other 32%. The time constant of the decay (for those adequately fitted by a single exponential) increased with depolarization, from an average value of 15 ms at -80 mV to 60 ms at +40 mV. 5. A relationship was noted between IPSC amplitude and decay time constant; IPSCs with larger peak amplitudes had a slower decay. One possible explanation considered for this finding was that transmitter persists in the synaptic cleft and rebinds to the receptors, thus prolonging the decay of the IPSC. 6. Consistent with the above hypothesis was the observation that the decays of miniature IPSCs (examined under conditions of reduced transmitter release) were faster, showed less variability, and were all adequately described by a single exponential function. Furthermore, the decay times were independent of the membrane potential, suggesting that the kinetic parameters of the GABA channel which shape the decay of these miniature IPSCs are independent of voltage. 7. Single-channel activity underlying whole-cell GABA responses could be recorded in isolated outside-out and inside-out patches of membrane. In isotonic choline chloride, single-channel amplitudes were linearly related to voltage and reversed at -1.8 ± 11.0 mV (n = 12). Under these conditions, the channel had a main conductance state of 20.8 ± 3.4 pS (n = 12). Transitions were observed from this main conductance state to other conductance states, e.g., two subconductance stages of 6 and 12 pS and one supraconductance state of 30 pS. 8. The linearity of the single-channel current-voltage (I-V) relationship suggests that the nonlinearity of the whole-cell I-V relationship results from a voltage-dependence of the kinetics of the GABA channel. This voltage dependence was quantified by calculating the number of channels open at the peak of I(GABA) for the different holding potentials. This value increased e-fold for every 80 mV of depolarization.