Abstract *Sridharan Ramaratnam and †Murthy B. Narasimha *Department of Neurology, Apollo Hospitals, Chennai ; and †National Institute of Epidemiology, Indian Council of Medical Research, Chennai, India . Purpose: To estimate the prevalence of epilepsy in India by meta-analysis of previously published and unpublished studies and determine the pattern of epilepsy by using community-based studies. Methods: As many as possible previously published and unpublished studies were identified on the prevalence of epilepsy from various parts of India by using multiple search strategies. The studies were assessed regarding the methods and definitions (population structure, method of case ascertainment, sensitivity and specificity of screening instrument, whether validation resurvey was done, definitions of epilepsy and active epilepsy, and inclusion of febrile fits/single seizures). The prevalence rates for rural and urban populations, as well as male and female ratios with 95% confidence intervals were calculated. The studies, which provided details on the age structure, age-specific rates, and patterns of epilepsy, were chosen for meta-analysis. The crude and age-standardized prevalence rates after accounting for heterogeneity, as well as pooled estimates of standardized prevalence, were calculated. The χ2 test was used for homogeneity testing. Heterogeneity correction was done by using a simple random-effects model. Results: We identified 21 studies conducted over the last 38 years (10 on urban and 14 on rural populations covering 11 states and two union territories) involving a population sample of 837,039, among whom 4,220 had epilepsy, giving a crude prevalence of 5.04/1,000. The nonstandardized crude prevalence per thousand (95% confidence interval) after correction for heterogeneity was 5.44 (3.79–7.09). Age-standardized rates (seven studies from seven states involving a population sample of 448,109), after correction for the variability in estimates of heterogeneity, revealed that the prevalence rate per 1,000 (95% confidence interval) was overall, 5.39 (3.78–7.01); males, 5.76 (4.23–7.28); females, 4.68 (2.94–6.42); urban, 6.02 (3.98–8.05); and rural, 4.77 (2.36–7.18). Although urban male and female subjects had higher prevalence compared with rural male and female subjects, the difference was not statistically significant. Age-specific prevalence rates were higher in the younger age group. The onset of epilepsy was mostly in the first three decades. The treatment gap was >70% in the rural areas. Generalized seizures constituted 45.45–86%, and partial seizures, 11.45–54.54%; 5.2–24.4% had a family history of epilepsy. The frequency of mental retardation ranged from 4.4 to 22.9%. Conclusions: The heterogeneity found among the various studies may be partly due to methodologic differences referred to earlier. The heterogeneity also may be due to prevalence of local types of epilepsy (hot-water epilepsy); differences in prevalence of cysticercosis, malaria, Japanese encephalitis, and head injuries; or due to denial of epilepsy on account of social stigma. The prevalence rates (crude and age-standardized) found in this meta-analysis are comparable to those in the Western countries and China. The estimated number of persons with epilepsy in India as per the census population for 2001 (1,025 million) would be 5.5 million. Because the rural population constitutes 72% of the Indian population, the number of persons with epilepsy in rural areas will be ∼3.6 million, among whom three fourths will not be receiving any specific treatment as per the present standards. The model used in this study is useful for estimating disease burden in large countries.