In families with febrile seizures and temporal lobe epilepsy, mutations affecting different GABAergic mechanisms suggest that failure of chloride conductance to limit depolarization may be directly epileptogenic. This “GABAergic disinhibition” hypothesis has been discounted historically for two reasons. First, early attempts to produce hippocampal sclerosis and epilepsy simply by eliminating hippocampal GABA neurons consistently failed to do so. Second, the notion persists that because clinical epilepsy diagnosis is typically delayed for years or decades after brain injury, temporal lobe epileptogenesis should be presumed to involve a complex pathological transformation process that reaches completion during this “latent period.” Recent advances clarify both issues. Although spatially limited hippocampal GABA neuron ablation causes only submaximal granule cell hyperexcitability, more spatially extensive ablation maximizes granule cell hyperexcitability and triggers nonconvulsive granule cell status epilepticus, hippocampal sclerosis, and epilepsy. Recent studies also show that disinhibited granule cells begin to generate clinically subtle seizures immediately post-injury, and these seizures then gradually increase in duration to become clinically obvious. Therefore, rather than being a seizure-free “gestational” state of potentially interruptible epileptogenesis, the “latent period” is more likely an active epileptic state when barriers to seizure spread and clinical expression are gradually overcome by a kindling process. The likelihood that an epileptic brain state exists long before clinical diagnosis has significant implications for anti-epileptogenesis studies. The location, magnitude, and spatial extent of inherited, autoimmune, and injury-induced disinhibition may determine the latency to clinical diagnosis and establish the continuum between the benign, treatable, and refractory forms of temporal lobe epilepsy.