The conventional methods of risk assessment were reviewed as an analytic framework for research. Studies have suggested that even under the best of circumstances such assessments pose serious problems for worker protection. Those who assess risk have argued that this assessment should be done separately from risk management so that risk assessment should not determine policy choices, but should serve only as an objective input to the policy process. They also have encouraged the use of risk assessment by decision makers to assist them in understanding the consequences of their choices. The decision of how to do the risk assessment should be guided by choices about risk management. In any risk assessment there are enormous uncertainties of a variety of types such as the toxic effects of a substance, the patterns of exposure, and the dose response relation. The burden of proof issue was considered to be complex. While it may be possible to state that there is no evidence that some factor is a risk, the question of whether there is any evidence to suggest that the factor is not a risk should also be asked. Also of importance has been the question of choice. When a worker chooses to remain on a particular job, knowing that there are risks involved, some have viewed that choice as an acceptance of the risk. The author suggests that the regulatory process must be made responsive to the issues presented, particularly the distribution of risk and the social effects of quantitative analysis.