319 patients with impulsive petit mal and 390 family members of the probands were examined clinically, genetically and electroencephalographically. The results are as follows. 1. The incidence of probands with genetic predisposition was 27.3%, which was higher than that (9.9%) found in 466 non-selected epileptic patients at the same clinic. The incidence was higher in female than in male probands (33.5%:21%). 2. The incidence of epilepsy among near relatives was 4.1% and was higher among female than male near relatives (5.4%:2.9%), higher among mothers than fathers (4.7%:1.9%), and higher among daughters than sons (9.8%:1.8%), but varied only slightly between sisters and brothers (4.6%:4.2%). The incidence was highest in sons and daughters (5.1%), intermediate in siblings (4.4%) and lowest in parents (3.3%). The relatives of female probands showed a higher incidence than those of male probands (6.0%:2.0%). 3. Impulsive petit mal was found in 15% of the 116 epileptic relatives, awakening grand mal in 17% and absence in 14%. 4. EEG examination showed unspecific paroxysmal abnormalities in 40% and specific abnormalities in 15% of 390 family members, incidences significantly higher than those in the control subjects. Relatives of female probands showed a higher frequency of paroxysmal EEG abnormalities than that in male probands (61%:47%). Offspring were found to have the highest, siblings an intermediate and parents the lowest frequency of paroxysmal as well as specific EEG abnormalities. 5. Of 1618 near relatives 118 (7.3%) showed epileptic seizures and/or specific EEG abnormalities. The incidence was higher in female than in male relatives (9.3%:5.5%) and higher in relatives of female probands than those of male probands (9.3%:5.2%). Offspring showed the highest (13.8%, 17% in daughters and 12% in sons), siblings an intermediate (6.7%, 8% in sisters and 5% in brothers), and parents the lowest level (5.2%, 8% in mothers and 3% in fathers). 6. Correlation between the incidence of EEG abnormalities and age among relatives was analyzed. 7. Pedigree studies showed maternal preponderance. 8. No proband was found to be born of a consanguineous marriage. There were many sibship cases in which one parent at least was found to have specific or paroxysmal EEG abnormalities. 9. It was concluded that the mode of inheritance of impulsive petit mal is most probably polygenic. Our hypothesis is (1) the etiology of the illness is more genetic less environmental in female patients (greater heritability). (2) Female subjects have a lower threshold for manifestation. 10. In our sample, 55% of male and 65% of female patients were married. Both figures were lower than those (78% for males and 77% for females) in the general population. 11. The importance of family study is emphasized.