We present the conclusions of two prospective studies of patients examined at their first manifestation of Graves' disease and treated with antithyroid drugs (ATD). The purpose of the first study was to investigate the effects of long-term treatment: the patients were given carbimazole in degressive doses without hormone replacement for 18 months, the followed up for 2 to 6 years after drug withdrawal. The second study was designed to determine the effect of treatment duration on the prognosis: the patients were given an ATD according to the same protocol for a duration randomly set at either 6 or 18 months, then seen again 2 years after ATD withdrawal. The results showed that after 18 months of treatment at least 50 percent of the patients could be expected to remain in remission for 6 years. Remissions were less frequent when treatment was shorter (41.7 percent after the 6 month treatment versus 61.8 percent after the 18 month treatment, with a 2 years' follow-up; P < 0.0.5). The relapses that occurred came early: 70 percent of them took place within the first post-treatment month. This article also provides evidence of high T3 and/or T4 levels without signs of thyrotoxicosis during the post-treatment clinical course; these exclusively biochemical relapses spontaneously disappeared and may have been expressing epidoses of active thyroiditis. Taking advantage of this study, we propose a treatment of varying duration according to the patient's status, and we underscore the value of pursuing studies in two preferential directions: (1) to try to assess the remission and evolution criteria of the disease (According to current knowledge, an end-treatment TSAb level above 300 percent indicates severity and supports the pursuance of ATD therapy); (2) to evaluate the prognostic implications ot ATD dosage, a treatment based upon high-dose ATD administration being likely to improve the prognosis.