Some of the many factors that influence dose selection in first-time-in-human studies are examined. These include animal toxicology, toxicokinetics, allometric scaling, pharmacokinetics, body surface area correlations, and integration of preclinical pharmacologic and toxicologic data. Appropriate preclinical evaluation and analysis may reduce the frequency and severity of unexpected toxic events arising during single-dose, phase I testing. However, significant intrinsic uncertainties in this process presently exist and will continue to exist well into the foreseeable future. With our present state of knowledge, we cannot provide a realistic and reasonable algorithm for ascertaining first-time-in-human doses: any decision tree would be too unwieldy. There are several rules of thumb that do have a place in the evaluation and decision-making process, however.