To examine whether lipid abnormalities contributed to the endemic peripheral vascular disease (PVD) in villages where arseniasis was hyperendemic in Taiwan, the authors studied 533 adults with Doppler ultrasound and lipid profiles including total cholesterol, triglyceride, high- and low-density lipoprotein cholesterol, apolipoprotein AI, and apolipoprotein B. Among them, 63 had PVD based on an ankle-brachial index < 0.90. Long-term arsenic exposure indices including cumulative arsenic exposure in mg/L-years, duration of drinking artesian well water in years, and duration of living in arseniasis-hyperendemic villages in years were calculated from detailed history obtained through standardized interviews based on a structured questionnaire and arsenic concentration in well water. Possible confounders including age, sex, body mass index, cigarette smoking, and disease status of diabetes mellitus and hypertension were considered in the analyses. None of the lipid profiles differed significantly between the presence and absence of PVD. The odds ratios for PVD did not differ among different quintiles of lipid profiles with the lowest quintile as the referent. However, a significant dose-response relation was found for the long-term arsenic exposure indices. The multivariate-adjusted odds ratios for cumulative arsenic exposure of 0.1 approximately 19.9 and > or = 20 mg/L-years were 2.77 and 4.68, respectively, compared with the unexposed. These results suggest that the PVD in arseniasis-hyperendemic villages is correlated with ingested inorganic arsenic and not with the lipid profiles.