The history of approaches to evaluating the hazards and risks of chemicals is briefly reviewed. The role of default options (generic approaches based on general knowledge in the absence of specific knowledge to the contrary) is discussed as a part of the risk assessment paradigm advanced by the National Academy of Science/National Research Council in 1983 and 1994. Examples are given of the impact of acquiring specific science to replace default options. An argument is made for developing specific science that would reduce uncertainty in risk assessments. Research on specific science would be guided by identified sources of uncertainty in the risk assessment process. The importance of using a research strategy that builds on human data is emphasized for validating new molecular and cellular biological assessment methods. The paper closes with a discussion of the tension between a hazard-based approach versus quantitative risk assessment in guiding risk management decisions. The former requires limited data, is qualitative, and easy to communicate, while the latter requires substantial data and is difficult to communicate. However, quantitative risk assessment provides a more rational basis for decisions on the allocation of both public and private resources for actions that will effectively minimize overall health risks to the public.