This article examines how scientists and regulators distribute the benefit of the doubt about drug safety under conditions of scientific uncertainty. The focus of the empirical research is the regulatory controversy over the hepatorenal toxicity of benoxaprofen in the United Kingdom and the United States. By scrutinizing the technical coherence of the arguments put forward by industrial and government scientists, it is concluded that these scientists are willing to award the commercial interests of the pharmaceutical industry an enormous benefit of the sccentific doubt, which is not consistent with the best interests of patients. Interpretative flexibility, the burden of proof falling on regulators and their trust in, and dependence on, industrial scientists facilitates that distribution of the benefit of the scientific doubt. However, regulatory authorities' need for viability and the rationality common to opposing scientific views suggest that it is possible, in principle, to alter this dominant trend. To achieve adequate patient protection, drug regulation in the United Kingdom and the United States requires extensive reform. Some preliminary policy changes are sketched.