Abstract Scientific debate surrounds the regulatory approach for evaluating carcinogenic risk of arsenic compounds. The arsenic ambient water quality criteria (AWQC), based on the assumption of a linear mode of action for skin cancer risk, results in an allowable limit of 0.018ppb in ambient waters; the drinking water Maximum Contaminant Level (MCL) was determined using a similar linear approach. Integration of results from recent studies investigating arsenic’s mode of action provide the basis for a change in the approach for conducting an arsenic cancer risk assessment. Results provide support for a concentration demonstrating a dose-dependent transition in response from those representing adaptive changes to those that may be key events in the development of cancer endpoints. While additional information is needed, integration of current research results provides insight for a new quantitative cancer risk assessment methodology as an alternative toxicologically-based dose response (BBDR) cancer modeling. Integration of the new experimental results, combined with epidemiological evidence, support a dose-dependent transition concentration of approximately 0.1μM arsenic. Some uncertainties remain; additional information from chronic in vitro studies underway is needed. Results to date also provide initial insight into variability in population response at low arsenic exposures.