Medical decision making can be viewed along a spectrum, with categorical (or deterministic) reasoning at one extreme and probabilistic (or evidential) reasoning at the other. In this paper we examine the flowchart as the prototype of categorical reasoning and decision analysis as the prototype of probabilistic reasoning. Within this context we compare PIP, INTERNIST, CASNET, and MYCIN—four of the present programs which apply the techniques of artificial intelligence to medicine. Although these systems can exhibit impressive expert-like behavior, we believe that none of them is yet capable of truly expert reasoning. We suggest that a program which can demonstrate expertise in the area of medical consultation will have to use a judicious combination of categorical and probabilistic reasoning—the former to establish a sufficiently narrow context and the latter to make comparisons among hypotheses and eventually to recommend therapy.