Glutathione (GSH) homeostasis in plants is essential for cellular redox control and efficient responses to abiotic and biotic stress. Compartmentation of the GSH biosynthetic pathway is a unique feature of plants. The first enzyme, gamma-glutamate cysteine ligase (GSH1), responsible for synthesis of gamma-glutamylcysteine (gamma-EC), is, in Arabidopsis, exclusively located in the plastids, whereas the second enzyme, glutathione synthetase (GSH2), is located in both plastids and cytosol. In Arabidopsis, gsh2 insertion mutants have a seedling lethal phenotype in contrast to the embryo lethal phenotype of gsh1 null mutants. This difference in phenotype may be due to partial replacement of GSH functions by gamma-EC, which in gsh2 mutants hyperaccumulates to levels 5000-fold that in the wild type and 200-fold wild-type levels of GSH. In situ labelling of thiols with bimane and confocal imaging in combination with HPLC analysis showed high concentrations of gamma-EC in the cytosol. Feedback inhibition of Brassica juncea plastidic GSH1 by gamma-EC in vitro strongly suggests export of gamma-EC as functional explanation for hyperaccumulation. Complementation of gsh2 mutants with the cytosol-specific GSH2 gave rise to phenotypically wild-type transgenic plants. These results support the conclusion that cytosolic synthesis of GSH is sufficient for plant growth. The transgenic lines further show that, consistent with the exclusive plastidic localization of GSH1, gamma-EC is exported from the plastids to supply the cytosol with the immediate precursor for GSH biosynthesis, and that there can be efficient re-import of GSH into the plastids to allow effective control of GSH biosynthesis through feedback inhibition of GSH1.