The thymus produces many more cells than it releases into the periphery. According to generally accepted models of T cell development most of this loss occurs in the thymic cortex, among CD4+8+ thymocytes. An interesting situation arises in the case of T cell receptor (TcR) transgenic mice in which all cells can potentially be positively selected, leading to a theoretical increase of about 30-fold in the survival rate of CD4+8+ cells and in their transition to mature CD4+8- or CD4-8+ thymocytes. This in turn should lead to a 30-fold increase in the size of the thymic medulla, in the emigration rate and in the size of the peripheral T cell pool. Increases in medullary or peripheral pool sizes of this magnitude are not seen in TcR transgenic mice. The question was therefore asked whether some form of homeostatic process regulated the size of the mature T cell pool and at what level it might operate. In this report we demonstrate that the increased rate of double-positive to single-positive transition in the TcR transgenic mice is directly reflected in an increased emigration rate, and that the medulla seems to be relatively efficient regardless of the number of cells passing through it. However, the potential increases in emigrant numbers in TcR transgenic mice are offset by the reduced size of the CD4+8+ thymocyte pool. It would appear then that regulation of T cell production, if it occurs, probably does so through regulation of the size of the CD4+8+ thymocyte pool. Mechanisms for regulation of this kind are not yet known.