Concerns about the quality and reproducibility of scientific papers have emphasised the uncertainty of observations in the biological sciences. Focussing on this problem in the types of science that often underlie regulations affecting population exposures has clear implications for regulatory studies. Recent comments on the IARC process of hazard identification in the field of carcinogenesis have identified problems in dealing with public information content and regulatory value of the Monograph system. What is the value of a system that does not clearly identify its methodology for defining a hazard and which does not provide realistic estimates of risk?