A new theory of the development of expertise in medicine is outlined. Contrary to existing views, this theory assumes that expertise is not so much a matter of superior reasoning skills or in-depth knowledge of pathophysiological states as it is based on cognitive structures that describe the features of prototypical or even actual patients. These cognitive structures, referred to as "illness scripts," contain relatively little knowledge about pathophysiological causes of symptoms and complaints but a wealth of clinically relevant information about disease, its consequences, and the context under which illness develops. By contrast, intermediate-level students without clinical experience typically use pathophysiological, causal models of disease when solving problems. The authors review evidence supporting the theory and discuss its implications for the understanding of five phenomena extensively documented in the clinical-reasoning literature: (1) content specificity in diagnostic performance; (2) typical differences in data-gathering techniques between medical students and physicians; (3) difficulties involved in setting standards; (4) a decline in performance on certain measures of clinical reasoning with increasing expertise; and (5) a paradoxical association between errors and longer response times in visual diagnosis.