In 1947, the Cornell psychiatrist Keeve Brodman and a handful of colleagues began developing what would become one of the most widely used health questionnaires of its time—the Cornell Medical Index (CMI). A rigidly standardized form, the CMI presented 195 yes-no questions designed to capture the health status of “the total patient.” Over the following decades, Brodman’s project of standardizing medical history taking gradually evolved into a project of mathematizing and computerizing diagnosis: out of the CMI grew the Medical Data Screen (MDS), an early computerized method of deriving diagnoses from patient data. At the same time, Brodman was beginning to work towards the MDS, another research team, headed by the television pioneer Vladimir Zworykin, and was developing a computer program that they hoped would make accurate diagnoses in the field of hematology. Drawing on these and other case studies, this project examines early efforts to computerize medical diagnosis and decision making. It explores how these efforts built upon, interacted with, and produced certain professional tensions, economic interests, disease constructions, cultural ideals, and material practices. Debates about how and whether to computerize diagnosis were often animated by larger debates about the nature of diagnosis and medical reasoning, the definitions of disease, and the authority as well as the identity of physicians and their patients.