Technological advances are rapidly leading toxicology and risk assessment to an age where there are advantages of routine use of biologically based modeling as an adjunct to laboratory experiments will be compelling. The biologically based model, a formal representation of the system being studied, serves both the experimentalist and the risk assessor. Benefits of its development include identification of data gaps, optimization of experimental design and, with sufficient validation, quantitative extrapolations between routes of exposure, species and doses. This article focuses on the role of biologically based modeling in support of the dose–response assessment component of risk assessment. The problem of uncertainty in dose–response assessment and the role of mechanistic studies combined with modeling to efficiently reduce uncertainty are discussed. The need for close cooperation between experimentalists, modelers, and risk assessors to achieve this efficiency is emphasized. A model for formaldehyde is briefly described to illustrate key issues.