We report an attempt to quantitate the relative contributions of the history, physical examination, and laboratory investigation in making medical diagnoses. In this prospective study of 80 medical outpatients with new or previously undiagnosed conditions, internists were asked to list their differential diagnoses and to estimate their confidence in each diagnostic possibility after the history, after the physical examination, and after the laboratory investigation. In 61 patients (76%), the history led to the final diagnosis. The physical examination led to the diagnosis in 10 patients (122%), and the laboratory investigation led to the diagnosis in 9 patients (1 10%). The internists' confidence in the correct diagnosis increased from 7.1 on a scale of 1 to 10 after the history to 8.2 after the physical examination and 9.3 after the laboratory investigation. These data support the concept that most diagnoses are made from the medical history. The results of physical examination and the laboratory investigation led to fewer diagnoses, but they were instrumental in excluding certain diagnostic possibilities and in increasing the physicians' confidence in their diagnoses. Ari-iving at most medical diagnoses requires information obtained from the history, the physical examination, and the laboratory investigation. In 1947 Platt claimed that in most cases the diagnosis can be made with the history alone.' In 1975 Hampton and co-workers attempted to evaluate the relative contributions of the history, the physical examination, and laboratory tests in making medical diagnoses in their study of 80 referral patients at a general medicine clinic.2 Internists were asked to record their presumptive diagnosis, their prediction of management, and their confidence in their diagnosis (or diagnoses) on a scale of 1 to 10 after reading the referral letter and taking the history, after completing the physical examination, and again after completing the laboratory investigation. These diagnoses were compared with those accepted two months after the initial visit. They found that the diagnosis predicted after taking the history agreed with the accepted diagnosis two months after the initial visit in 66 of the 80 patients (82%). The physical examination led to the diagnosis in 7 patients (9%), and the laboratory investigation led to the diagnosis in the other 7 patients (9%).In a similar study of 630 of his own patients at a clinic with a "cardiological bias," Sandler reported that 56%, 17%, and 23% of his diagnoses were made from the history, physical examination, and laboratory investigation, respectively.3 Gruppen and associates reported in a study of 119 patients at a primary care walk-in clinic that greater than 90% of their diagnoses were made after the physician heard the chief complaint, read the nurse's intake note, and completed the history taking.4 In 95% of their patients, the correct diagnosis appeared in the differential diagnosis list after the history was taken.During the 16 years since the initial study by Hampton an...