The participation of genetic factors in convulsive disorders is a controversial subject. Some investigators consider it to be high and others low. Autosomal dominant, autosomal recessive, sex-linked recessive, and polygenic mechanims of inheritance have been proposed 4 ' 5 ' 6 .One of the factors that brings special difficulty to define the influence of heredity is the heterogeneous nature of the problem. In an attempt to isolate a supposedly more homogeneous group of patients, and to try to define wheter or not a genetic predisposition is present, we developed a study of subjects who had had febrile convulsions. Febrile convulsions was defined as an epileptic event ocurring in a previously normal child during a febrile illness unrelated to a nervous system disease 1 ' 3 ' 7 . MATERIAL, METHODS AND RESULTS Patients attending the Hospital das Clinicas, Faculdade de Medicina de RibeirãoPreto, USP, who had shown at least one well-characterized episode of febrile convulsion were referred to the Febrile Convulsions Ambulatory, and studied during the period of January to December, 1979. A complete family history was obtained in addition to clinical and laboratory data. Relatives with febrile convulsions were considered only when the report given left no margin for doubt. Of the 128 patients studied, 18 (14,3 % of the sample) had at least one relative affected by febrile convulsions.The pedigrees of these patients can be seen in figure 1. The age of the probands variei between 8 months and 16 years, with a mean age of 4 years and 6 months.Family recurrence was always unilateral (only on the paternal or only on the maternal side), which is not in agreement with a polygenic mechanism of inheritance 8.