We undertook a community-based case-control study on persons with active epilepsy residing in Kerala, southern India. Using a standardized questionnaire, we collected information from 362 cases and 362 controls. In the final multivariate model, family history of epilepsy (odds ratio=7.8, 95% confidence interval=3.2-18.8, P=0.000), antecedent history of febrile seizures (7.7, 4.3-14.0, 0.000), birth by complicated delivery (6.8, 2.1-21.8, 0.001), and neonatal seizures (7.8, 1.7-35.4, 008) emerged as strong independent predictors of epilepsy, followed in decreasing order by mental retardation, prematurity, maternal age 30, perinatal distress, and incomplete immunization. There were more similarities than differences in the distribution of risk factors between generalized and localization-related epilepsy syndromes. Our findings suggest interplay between genetic and acquired factors in the pathogenesis of epilepsies, and underscore the need for improvement in obstetric and neonatal care to minimize the epilepsy burden in low-income countries.