Ketogenic diets (KDs) are designed to create the metabolic conditions of fasting, which was among the earliest therapies discovered for epilepsy. The major measures used to evaluate dietary effectiveness have been the levels of urinary ketone bodies and the successful reduction of seizure activity. Modifications of the "classical" animal fat KD have been used in an attempt to boost ketonuria or ketonemia, increase palatability and compliance, and reduce side effects. Studies of KDs in experimental animals have been largely confined to rodents (mice and rats) for reasons of cost and convenience, and both have been found to be protected against experimentally induced seizures following consumption of KDs. Most of these studies have been designed to test hypotheses about the mechanism(s) by which reductions in carbohydrate or increases in fat result in elevated seizure threshold, decreased seizure duration, and decreased seizure severity. So far, underlying mechanisms have proven elusive. Rodent studies have led to a degree of general agreement that ketone levels per se do not correlate well with seizure protection, that reduction of glucose levels is fundamentally important, and that calorie restriction is additive to high fat diets in providing seizure protection.