The human thymus is required for establishment of the T cell pool in fetal life, but postnatal thymectomy does not lead to immunodeficiency in humans. Because thymectomy in humans is performed for treatment of myasthenia gravis (MG), we have studied patients with MG for effects of thymectomy on peripheral blood (PB) naive (CD45RA(+), CD62L(+)) and memory (CD45RO(+)) T cells. We have also determined the effect of thymectomy on levels of PB cells containing signal joint TCR delta excision circles (TRECs), a molecular marker of thymus emigrants that have divided few times after leaving the thymus. In 17 nonthymectomized and 26 thymectomized MG patients studied at varying times after thymectomy (1 day to 41 years), we found no significant mean difference in PB T cell TREC levels between ages 40 and 80 years. However, both thymectomized and nonthymectomized MG patients had lower PB T cell TREC levels than did age-matched normal subjects (p < 0.0001 for both). These data demonstrated that MG itself or treatment for MG decreased thymopoiesis independent of thymectomy. Next, to control for disease activity and treatment, we prospectively studied 10 MG patients before and from 27 to 517 days after thymectomy. We found that thymectomy decreased CD4 or CD8 T cell TREC concentrations most when thymopoiesis was active before thymectomy (six of six patients), but had little effect in patients when thymopoiesis was minimal (four of four patients). In contrast, there was no significant effect of thymectomy on absolute numbers of naive PB T cells. Thus, in MG, removal of a thymus with active thymopoiesis resulted in a significant fall in PB TREC(+) T cells postthymectomy.