AIMS/HYPOTHESIS: Beta cell development is sensitive to glucocorticoid levels. Although direct effects of glucocorticoids on pancreatic precursors have been shown to control beta cell mass expansion, indirect effects of these hormones on pancreatic development remain unexplored. This issue was addressed in mice lacking the glucocorticoid receptor (GR) in the whole organism. MATERIALS AND METHODS: The pancreatic phenotype of GR(null/null) mice was studied at fetal ages (embryonic day [E]) E15.5 and E18 by immunohistochemistry and beta cell fraction measurements. To distinguish between direct and indirect effects, mutant E15.5 fetal pancreata were grafted under the kidney capsule of immunodeficient mice and analysed after 1 week. RESULTS: E18 GR(null/null) fetuses had smaller digestive tracts and tiny pancreata. Massive pancreatic disorganisation and apoptosis were observed despite the presence of all cell types. E15.5 GR(null/null) mutants were indistinguishable from wild-type regarding pancreatic size, tissue structure and organisation, beta cell fraction and production of exocrine transcription factor Ptf1a, neurogenin 3 and Pdx-1. Grafting E15.5 GR(null/null) pancreata into a GR-expressing environment rescued the increased apoptosis and mature islets were observed, suggesting that GR(null/null) pancreatic cell death can be attributed to indirect effects of glucocorticoids on this tissue. Heterozygous GR(+/null) mutants with reduced GR numbers showed no apoptosis but increased beta cell fraction at E18 and the adult age, strengthening the importance of an accurate GR dosage on beta cell mass expansion. CONCLUSIONS/INTERPRETATION: Our results provide evidence for GR involvement in pancreatic tissue organisation and survival through indirect effects. GR does not appear necessary for early phases, but its accurate dosage is critical to modulate beta cell mass expansion at later fetal stages, presumably through direct effects.