This paper is concerned with the policy-making processes and the nature of the information available to those who participate in them. At a very simple level of concern is the issue of how to define what is a risk to which governments should pay attention in the public interest. For those responsible for making policy decisions there are real dilemmas. The policy process itself is inadequate to deal with the processing of scientific information about risk. A representative parliamentary system is notoriously ill-equipped to cope with a multiplicity of information sources. The question of which particular scientific voice should be regarded as legitimate is problematic. There is no single institutional centre to identify particular problems. It is this latter question that creates the most difficulty for policymakers since they need to base their justification of policy on the most valid of grounds. In the past objective scientific fact has been so regarded. Today there are as many different scientific opinions as there are advisers. Problems are defined and redefined depending on the control of the agenda at any one point in time. Some suggestions are made about how these dilemmas might be addressed.