The objective of this study was to determine if cytogenetic markers can be used as indicators of prior exposure to arsenic compounds. Baseline sister chromatid exchange (SCE) and mitomycin C-induced (MMC) SCE were measured in four study populations recruited from a blackfoot (BF) disease endemic area, including 22 patients with cancer (CA) only, 8 patients with both BF disease and CA (BF+CA), 10 patients with BF disease only, and 26 healthy residents (HRs). Another group of 23 healthy, nonarsenic-exposed workers were recruited as external healthy controls (HCs). Characteristics of study population were collected by questionnaire, and 10 ml of venous blood were drawn for lymphocyte culture. The results showed that the frequencies of baseline SCE did not differ among the five study groups. The frequencies of delta SCE (MMC-induced SCE minus baseline SCE) in CA only, BF disease only, and HRs, three arsenic-exposed groups, were significantly higher than in HCs. The frequency of delta SCE in the BF+CA group was nonsignificantly higher than in HCs, probably due to small sample size. The frequencies of both baseline SCE and delta SCE did not differ among CA only, BF disease only, BF+CA, and HR groups. The observation that baseline SCE did not increase in the arsenic-exposed populations indicates either that exposures were insufficiently high to change this marker or that lesions did not persist. The increased SCE response to MMC in arsenic-exposed populations suggests that previous arsenic exposure may result in hypersensitivity of human lymphocytes to carcinogens and/or mutagens. Both baseline SCE and delta SCE were not different among patients with arsenic-induced diseases and healthy normal residents, indicating that hypersensitivity may have been due to previous arsenic exposure but was not associated with disease status.