In the absence of a prospective randomized study of patients treated conservatively or with thymectomy, a computer-assisted retrospective matched study was devised. Of 563 patients treated for myasthenia gravis without thymoma up to 1965, 104 had thymectomy. With computer assistance, each surgical patient was matched with a medical patient on the basis of age, sex, and severity and duration of disease. On this basis 80 of the 104 surgical patients could be matched satisfactorily. There were 16 males and 64 females in each of the matched surgically treated and medical control groups. A complete remission was experienced by 27 of the 78 patients in the surgical group as compared to 6 of the medical group. Improvement was noted by 26 of 78 surgically treated patients and 13 of 78 receiving medical treatment. Survival curves show a significantly better chance of long-term survival for patients having thymectomy. Thirty-four patients in the medical group had died as compared to 11 in the surgical group. Comparison of survival in relation to sex, duration of symptoms, or age (<30 or >30 years) did not show a significant difference. Until more effective treatment is available for myasthenia gravis, thymectomy deserves consideration for both sexes, and with increased age or long duration of symptoms.