Epilepsy is associated with excess mortality of two to three times in developed countries. Precise epidemiological data on mortality and cause of death are not available from India or most other developing countries. This study was carried out to estimate the mortality rates and to identify the demographic and clinical characteristics associated with mortality in a hospital based cohort of epilepsy patients. A cohort of patients enrolled in the epilepsy clinic in 1985 was followed up till 1997 (12 years). The mortality rate, demographic and clinical correlate of mortality were analysed for 246 patients (men 161, women 85) who had complete data. Cause of death was not examined in this study. Eighteen (men 15, women 3) of the 246 patients (7.3%) had died during the follow up period of 12 years. The crude death rate for the state of Kerala for the year 1990 (mid period of the study) was 5.9 per thousand population. The demographic and clinical characteristics of those who died (corresponding figures for survivors are given in brackets) were as follows: mean age 33.6 years (22.8 years), presence of abnormal neurological examination 38.9% (15.4%), mental retardation 33.3% (12.8%), abnormal CT scan 38% (21.5%). Regarding the seizure frequency at the time of enrollment and eventual mortality, there were no deaths among patients who had an Engel's seizure score of less than or equal to 4 (no seizures or nocturnal seizures only). The mortality was 5% for an Engel's score of 5 or 6 and 11% for an Engel's score greater than 6. Within the group with epilepsy, higher seizure frequency at the time of initial evaluation was associated with excess mortality. Abnormality on neurological examination, older age group and male sex were other factors that correlated with excess mortality.