1. In whole cell patch-clamp studies on cultured rat embryonic spinal cord and medullary neurons bathed in tetrodotoxin, DL-2-amino-5-phosphonovaleric acid, and 6-cyano-7-nitroquinoxaline-2,3-dione, large and long-lasting spontaneous inhibitory postsynaptic currents were occasionally recorded. The amplitudes of these events were 1 order of magnitude larger than those of spontaneous miniature inhibitory postsynaptic currents. Because these large currents had reduced amplitudes in calcium-free saline and in solutions containing glycinergic or GABAergic antagonists, we conclude that they were probably produced by large and prolonged release of glycine and/or 4-amino-n-butyric acid (GABA), which subsequently bind to their postsynaptic receptors. 2. The frequency of spontaneous miniature postsynaptic currents increased dramatically during the long, slow decay phase of these large postsynaptic currents. Considering the requirement for extracellular calcium for the occurrence of these large responses, we hypothesize that this increased frequency reflected an increased intracellular calcium concentration in the presynaptic terminal. 3. Similar evidence for large inhibitory postsynaptic currents and prolonged transmitter release was observed in cell-attached patches, which also exhibited the smaller, spontaneous miniature inhibitory postsynaptic currents, suggesting that these large events are properties of single synaptic terminals. 4. A comparison of the properties of these large inhibitory postsynaptic currents recorded in whole cell mode or cell-attached patches showed no statistically significant differences. The overall mean values, then, are 13.9 +/- 1.6 (SE) ms and 4.5 +/- 0.5 s for the 10-90% rise time and duration, respectively. Furthermore, these large events had amplitudes that were 11-fold larger than the mean amplitude of the miniatures (i.e., mean amplitude ratio of 10.8 +/- 0.5). 5. Periodic large increases in the frequency of spontaneous miniature inhibitory postsynaptic currents occurred in both cell-attached patches and in the whole cell mode, and these increases were only sometimes associated with the large inhibitory postsynaptic currents. The rhythmicity in both recording configurations had similar temporal characteristics, with average interburst intervals of 5 and 12-14 s. Presumably these bursts of spontaneous miniature postsynaptic currents reflected periodic oscillations in the Ca2+ concentration in presynaptic terminals. 6. Both the probability and the frequency of occurrence of large inhibitory postsynaptic currents doubled during the 7-day period of time in culture when experiments were performed, suggesting that these large currents may play a role during development.