Ninety patients with Graves' disease were given treatment with antithyroid drugs. The first course, which is almost always immediately successful, gave in the long term 46 failures and 44 successes. Among the latter, 38 had a follow up equal or greater than one year after interruption of the course. Among the 78 cases sufficiently well followed up for the whole period of treatment (of which 21 had at least two courses of treatment), we noted 23 failures and 55 successes (70%) for which the average follow up was 61 months. If one requires a follow up equal or greater than one year after the end of treatment, we noted 47 good results out of 70 fully documented cases, i. e. 67%. The possible failures almost always occurred less than one year after the end of treatment. The delay since the last interruption of antithyroid drugs is thus of prognostic value. The result of treatment was not found to be related to any of the clinical or laboratory parameters found during the initial stage of the disease, nor with the type of antithyroid drug used. The addition of thyroid hormone to antithyroid drugs does not seem to be of any use. On the other hand, delay in starting treatment seems to favour the failure of treatment. Other interesting prognostic factors may be sought in the kinetics of TSH, the T3-suppressibility, the early radioiodine uptake, the estimation of thyrostimulating antibodies, and the study of HLA histocompatibility antigens.