MODELING OF HUMAN DISEASE in animals forms the basis of much of biomedical research. Animal models have been used to explore normal physiology, the pathophysiology of human disorders, and to test the safety and efficacy of new therapies in preclinical studies. In the field of lung research, much of our understanding of lung physiology under normal and pathological conditions derives from classic studies in dogs and sheep, including studies of normal pulmonary function by Julius Comroe (6), studies of the distribution of ventilation and perfusion by James West (25), studies of the pulmonary vasculature and alveolar surface tension by Solbert Permutt (10, 18), and studies of the mechanisms of formation of pulmonary edema by Norman Staub (23, 24) and the resolution of pulmonary edema by Michael Matthay (15). In lung-related biomedical research, animal models have been developed for most of the common human lung disorders including pneumonia, acute lung injury, asthma, chronic obstructive pulmonary disease, pulmonary fibrosis, and pulmonary hypertension. Although these models have been invaluable in improving the mechanistic understanding and testing potential treatment modalities for human lung disorders, all animal models have inherent limitations. For investigators, readers of the literature, and scientific reviewers of grants and manuscripts, an understanding of the pros and cons of animal models of human lung disease is critical. The need for every investigator to have a basic familiarity with animal models is illustrated by the fact that in the past six months, the majority (102 of 159) of the original articles published in The American