The most interesting recent development in the treatment of patients with epilepsy has been the apparent beneficial effect of measures associated with ketosis, viz., fasting and a high fat—low carbohydrate diet. During the past six years, we have been making observations on this subject with two points in mind: (1) to observe the clinical effect of fasting on seizures, and (2) to study the changes which take place in body metabolism that might explain the therapeutic results obtained. With reference to the latter consideration, certain changes in blood chemistry and metabolism which we believe to be without significance for epilepsy have been published.1 Our observations concerning the changes which are of apparent effect, in both fasting and fat diet and in other procedures designed to influence the frequency of seizures, have been summarized in a recent publication2 and will be presented in detail elsewhere.8 Our object in