Process alternatives, policy options and product life-cycles can be compared in terms of their potential contribution to toxicological effects associated with chronic exposures. These applications involve relative comparisons and can be performed using standard risk assessment paradigms. However, unlike most risk screening approaches that aim to be somewhat conservative, it is important that the comparisons are made using consistent data and assumptions. In this presentation, the uncertainty associated with the toxicological component of relative comparisons is considered and significant differences with the techniques used in risk screening are highlighted. It is demonstrated that the uncertainty associated with the toxicity component for each chemical in a relative comparison can be four orders of magnitude for aquatic ecosystems, for non-carcinogenic impacts in mammalian species including humans and for carcinogenic impacts. This uncertainty is dependent on the number of measured test results available, the test endpoint applicability (e.g. acute, chronic), the types of extrapolations performed and what estimation models, if any, are used.