The immunosuppressive effects of antithyroid drug therapy are well recognized; however, the cellular mechanisms underlying their action remain largely unknown. In the present paper we have prospectively analyzed the in vivo effects of methimazole treatment on a large number of circulating B cell subsets, involved in the effector phase of the immune response, in a group of 18 hyperthyroid patients with Graves' disease (GD). The patients were sequentially studied before (day 0) and 7, 14, 30, 90 and 180 days after methimazole therapy. The results were compared with both a group of 19 age- and sex-matched healthy controls and a group of 20 untreated/euthyroid GD patients in long-term remission. The combination of flow cytometry and three colour immunofluorescence revealed a clear increase (P < 0.001) in the numbers of circulating total B cells (CD19+) due to a significant increase (P < 0.001) in the CD5+, FMC7+, CD5+/ FMC7+ and CD23+ B cell subsets in hyperthyroid GD patients with respect to both healthy individuals and to GD patients in long-term remission. The absolute numbers of all these B cell subsets analyzed before treatment, although abnormal, were not statistically different from those observed during the whole period of therapy. When comparing the percentages of these B cell subsets during treatment, significant changes (P < 0.001) were only observed in the proportion of CD5+, CD5+/FMC7+ and CD5- B cells at the end of the follow-up period with respect to those found both before and during the first month of therapy. Whereas CD5+ and CD5+/FMC7+ B cells decreased (P < 0.001) after 3 months of therapy, CD5- B cells showed a significant increase (P < 0.001) at the end of therapy. It is remarkable that the percentage of CD5+, CD5+/FMC7+, CD5- and CD23+ B cell subsets were abnormal during the whole period of treatment and that they never reached normal values. These results show that, in vivo, GD patients treated with methimazole exhibited an abnormal but rather stable pattern of B cell distribution, similar to that present in hyperthyroid untreated GD patients, except for the CD5+ and CD5- B cell populations. Our findings suggest that in vivo methimazole therapy would not directly have an important influence on circulating B cell subsets.