It is now evident that prolonged febrile seizures in childhood, or an episode of status epilepticus at any age, can produce the highly characteristic pattern of hippocampal cell loss and shrinkage that is seen later in life, when patients develop temporal lobe epilepsy. Seizure-induced and presumably excitotoxic pathology includes neuronal loss, reactive gliosis, aberrant synaptic reorganization of surviving cells, and hippocampal tissue shrinkage that may alter extracellular space and affect ionic homeostasis. Whether any of these pathological effects of prolonged excitation play a causal role in the epileptogenic process that ultimately leads to spontaneous afebrile seizures remains a subject of intense interest. Two hypotheses have been suggested to explain how seizure-induced neuronal loss might initiate the epileptogenic process. One hypothesis suggests that normal inhibition and excitability is maintained by vulnerable non-principal cells, and that their loss deactivates inhibitory neurons, rendering principal cells disinhibited and hyperexcitable. The other hypothesis regards the initial loss as a stimulus for normally unconnected principal cells to form aberrant recurrent excitatory connections. Additional influences undoubtedly include a “kindling” process that gradually overcomes polysynaptic inhibition, and changes in extracellular space that may facilitate synaptic and ephaptic depolarization. Identification of the suspected substrates of epileptogenesis will serve as a stimulus for future progress and provide direction for new experimental designs.