In studies of focal epilepsy it is frequently assumed that the interictal spike is the elementary form of epileptic activity and that seizure, or ictal, episodes evolve by temporal summation and spatial expansion of interictal paroxysms. We examined this hypothesis in an in vitro model of acute focal epilepsy produced by perfusing rat hippocampal slices with solutions containing moderately elevated concentrations of K'. Some of the preparations treated in this way displayed recurring electrical seizures in the CA1 field. Each seizure episode typically evolved by a seemingly smooth progression of brief interictal bursts into sustained ictal discharge. However, exposure of preparations showing electrical seizures to blockers of synaptic transmission or to cholinergic agonists abolished interictal spiking in all hippocampal fields but did not impede the initiation of ictal episodes in area CA1. Likewise, severing the connections between areas CA3 and CA1 abolished interictal spiking in area CA1 without disrupting the initiation of seizures in this region. These data clearly show that in this model, focal seizures arise independent of interictal spikes and through different cellular mechanisms. While interictal electrogenesis requires chemical synaptic excitation, ictal episodes can be initiated and maintained by nonsynaptic neuronal interactions. Jensen MS, Yaari Y . The relationship between interictal and ictal paroxysms in an in vitro model of focal hippocampal epilepsy. Ann Neurol 1988;24:591-598 Two patterns of neuronal discharge are typically generated by epileptic cortex: interictal bursts responsible for the spikes in the electroencephalogram (EEG), and zctal paroxysms, or xezztlra { 1, 21. The interictal burst is a brief (approximately 50-100 msec) synchronous neuronal discharge that originates in a defined focus and is usually contained within it by synaptic and intrinsic inhibitory neuronal responses 131. Although it