Background: Arsenic drinking water concentrations >100 µg/L are a known cause of cancer, but the risks at lower concentrations are unknown and controversial. With its unusual geology and good information on past exposure, northern Chile is one of the best places in the world to investigate the long-term impacts of arsenic. Methods: We performed a lung cancer case-control study in areas in northern Chile with a wide range of drinking water arsenic concentrations. Previously we reported evidence of high lung cancer odds ratios (OR) for arsenic concentrations >100 µg/L. In this analysis, we calculated lung cancer ORs, adjusted for age, sex, and smoking, for those 92 cases and 288 population-based controls who had no known exposure above 100 µg/L at any point in their lives. Results: For exposures =40 years ago, lung cancer odds ratios (ORs) adjusted for age, sex, and smoking for increasing tertiles of arsenic exposure were 1.00, 1.43 (90% confidence interval, 0.82- 2.52), and 2.01 (1.14-3.52) (p-trend=0.02). Median arsenic water concentrations by tertile were 6.5, 23.0, and 58.6 µg/L. For subjects under age 65, corresponding ORs for exposures before age 25 were even higher. Adjustments for occupation, diet, and socioeconomic status had little impact on results. Conclusion: These findings provide rare new evidence that drinking water arsenic concentrations below 100 µg/L are associated with increased risks of lung cancer. These results also suggest that early-life exposure may cause marked increases in lung cancer in younger adults at arsenic water concentrations approaching those commonly found in the United States.