OBJECTIVE: Methods and results are presented for an arsenic exposure assessment integral to an epidemiological case-control study of arsenic and cancer-the European Commission funded ASHRAM (Arsenic Health Risk Assessment and Molecular Epidemiology) study carried out in some counties of Hungary, Romania and Slovakia. METHODS: The exposure history of each participant (N = 1,392) was constructed by taking into account how much water they consumed (as water, in drinks and in food), sources of drinking water in their various residences over their lifetime, and the concentrations of arsenic in their various water supplies measured by Hydride Generation-Atomic Absorption Spectrometry (HG-AAS). Concentrations of arsenic in previous water supplies were either derived from contemporary analyses of the same source, or from routine historical data from measurements performed by the authorities in each country. Using this approach, 80% of the recorded lifetime residential history was matched to an arsenic concentration. Seven indices of current, life time, and peak exposure were calculated. RESULTS: The exposure indices were all log-normally distributed and the mean and median lifetime average concentrations were in Hungary 14.7 and 13.3 microg l(-1), Romania 3.8 and 0.7 microg l(-1) and in Slovakia 1.9 and 0.8 microg l(-1), respectively. Overall 25% of the population had average concentrations over 10 microg l(-1) and 8% with exposure over 50 microg l(-1). CONCLUSIONS: Careful assessment of arsenic in drinking water supplies (both current and previous) enabled the majority of study participants' cumulative lifetime of potential exposure to arsenic in residential water to be characterised.