Exposure, as well as toxicity, determines whether rodenticides present real environmental hazards to nontarget animals. In order to combine exposure and toxicity, a compartment model is proposed which distinguishes transfer processes from accumulation of residues. The published literature relevant to the model is analysed, and some important gaps in knowledge are highlighted. Simple sub-models of rat feeding behaviour and mortality are combined into a simulation model which generates data on both efficacy of control and build-up of residues in live rats and carcases. The roles of feeding parameters (e.g., palatability, availability of alternative food) as well as toxicity are emphasised by the simulation results.