The drug development process is divided into phases with decisions required on compound selection and promotion to each subsequent development phase. In preclinical drug development the main objective is to bring the compound into human trials and there is an inability of many preclinical information packages to predict clinical responses. Since clinical responses are functions of the dose, the human dose anticipation should be a key deliverable of any preclinical package of drug candidate. The human dose should be anticipated by integration of information from multiple sources, in vitro and in vivo, non-human and human, using a variety of methodologies and approaches. Prediction of human safe and active dose relies on the availability of validated animal models for effect. Although there are many exceptions to the rule, the paper defines a four-step approach for the anticipation of human dose for first-in-man trials: 1, characterization of non-human exposure-response relationships; 2, correction for interspecies differences; 3, diagnosing compound absorption, distribution, metabolism and excretion (ADME) properties and prediction of human pharmacokinetics; and 4, prediction of human dose-responses and dose selection for phase I protocols.