Antiepileptic drug (AED) consumption, expressed according to a standardized measurement of the average daily doses of each active principle (the defined daily dose, DDD), was assessed as a source of morbidity data to calculate the prevalence of AED takers. In a population of 51,220 from three areas in Italy (Arcisate, Treviglio, S. Giovanni Rotondo) all the patients treated with AEDs, traced by the local general practitioners, were examined by a neurologist. The diagnosis of epilepsy was confirmed in 199 of 223 AED takers, giving an overall prevalence rate of 3.94 per 1,000 (Arcisate 3.96; Treviglio 4.04; S. Giovanni Rotondo 3.89). There was a significant overlap between the observed and expected number of AED takers (the latter obtained from regional sales and expressed in DDD). Prevalence rates calculated from drug sales were 5.58 in Arcisate and 6.11 in Treviglio. The ratio of patients with epilepsy to AED takers in the two areas was 0.7 and corresponded to the mean proportion of AEDs prescribed for epilepsy in the years 1983–1988 by a sample of Italian physicians. Although the prevalence rate calculated from AED sales tends to approximate that of treated epilepsy, AEDs can be reasonably used as ''tracers'' of disease for the easy access to drug sales in Italy, the appropriateness of the DDD as a standard measure of the daily doses of AED, the steady pattern of consumption, the comparability of data covering different areas and periods, and the stable proportion of drugs delivered for epilepsy.