1. Synaptic potentials induced by 4-aminopyridine (4-AP) were recorded intracellularly from rat neostriatal neurons in an in vitro slice preparation. EC50 for this 4-AP action was approximately 120 microM. The threshold for activation of synaptic potentials was 5 microM. 2. 4-AP-induced synaptic potentials appeared stochastically. Most were blocked by 1 microM tetrodotoxin or 400 microM Cd2+. Therefore they reflect a release of neurotransmitters dependent on both Ca2+ entry to the terminals and action potential firing. 3. Bicuculline (BIC) (< or = 10 microM), a gamma-aminobuturic acid-A (GABAA) antagonist, blocked about half of the 4-AP-induced synaptic potentials. This suggests that intrinsic inhibitory connections within the neostriatum are activated by 4-AP administration. 4. 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; < or = 10 microM) plus D-2-amino-5-phosphonovaleric acid (D-APV; < or = 100 microM) blocked most of the BIC-resistant 4-AP-induced synaptic potentials. This suggests that 4-AP induced release of glutamate (GLU) from extrinsic glutamatergic afferents. As most glutamatergic afferents are extrinsic, these afferents then would be able to fire spikes and release transmitter for several hours after they are cut from their somata. 5. If CNQX plus D-APV were administered before BIC, neostriatal neurons responded in different ways. In one half of the neurons, all induced synaptic potentials were blocked. This suggests that most GABAergic intrinsic connections between neostriatal neurons are activated indirectly by 4-AP. 4-AP would first activate extrinsic glutamatergic afferents and these in turn would activate GABAergic intrinsic neurons and connections. 6. In the remaining half of the recorded neurons, administration of CNQX plus D-APV blocked most, but not all of the 4-AP-induced synaptic potentials. The synaptic potentials that remained had a characteristic pattern: they were high amplitude, rhythmic, bursts of synaptic potentials. They were blocked by BIC (5 microM) but not by mecamylamine (> 10 microM). This suggests that these bursts of synaptic potentials were GABAergic and generated by intrinsic neurons. Therefore these neurons would not innervate all neostriatal neurons equally but just a subset of them. 7. Records from an identified aspiny neostriatal interneuron, obtained from the same preparation, are shown. This interneuron fired in bursts and its morphologically and physiologically similar to the recently described, fast spiking, parvalbumin immunoreactive, GABAergic, aspiny interneuron is functional in the slice preparation.(ABSTRACT TRUNCATED AT 400 WORDS)