This entire volume is directed atthe premises that (1) animals can serve as accurate predictive models of toxicity in humans (or other species), (2) the selection of an appropriate species to use is key to accurate prediction in man, and (3) understanding the strengths and weaknesses of any particular model is essential to understanding the relevance of speci‹c target organ toxicities to what would be expected in humans. Each of these premises requires some examination. Historic recognition of these facts (Priestly, 1772) has led to uses ranging from tasting food for poisons to the use of canaries in coal mines (Burrell, 1912).