Population-based studies about epilepsy have been informative about frequency, causes, and outcome of epilepsy. Because epilepsy is a heterogeneous disorder, studying populations of categories of patients is more likely to provide critical insights than is “lumping” all of epilepsy together. The 1981–1989 Classification scheme of the International League Against Epilepsy (ILAE) has been very useful for population-based studies but has some significant shortcomings. Studying a long list of specific epilepsy syndromes is not likely to be useful for population-based studies because the number of patients with each syndrome is so small and comparison with previous research would become very difficult. The author argues that with some significant modifications, the backbone structure of the 1989 Classification scheme could retain its usefulness for further research and clinical care.