This review summarizes the pathogenetic role of naturally occurring mutations of G protein genes in endocrine diseases. Although in vitro mutagenesis and transfection assays indicate that several G proteins have mitogenic potential, to date only two G proteins have been identi®ed which harbor naturally occurring mutations, Gsa, the activator of adenylyl cyclase and Gi2a, which is involved in several functions, including adenylyl cyclase inhibition and ion channel modulation. The gene encoding Gsa (GNAS1) may be altered by loss or gain of function mutations. Indeed, heterozygous inactivating germ line mutations in this gene cause pseudohypoparathyroidism type Ia, in which physical features of Albright hereditary osteodystrophy (AHO) are associated with resistance to several hormones, i.e. PTH, TSH and gonadotropins, that activate Gs-coupled receptors or pseudopseudohypoparathyroidism in which AHO is the only clinical manifestation. Evidence suggests that the variable and tissue-speci®c hormone resistance observed in PHP Ia may result from tissuespeci®c imprinting of the GNAS1 gene, although the Gsa knockout model only in part reproduces the human AHO phenotype. Activating somatic Gsa mutations leading to cell proliferation have been identi®ed in endocrine tumors constituted by cells in which cAMP is a mitogenic signal, i.e. GH-secreting pituitary adenomas, hyperfunctioning thyroid adenomas and Leydig cell tumors. When the same mutations occur very early in embryogenesis they cause McCune±Albright syndrome. Although these mutations would in principle confer growth advantage, studies failed to detect differences in the clinical and hormonal phenotypes, suggesting the existence of mechanisms able to counteract the activation of the cAMP pathway. Activating mutations of Gi2a have been identi®ed in a subset of ovarian, adrenal and pituitary tumors, but their prevalence and signi®cance are still controversial. Finally, although Ga subunits are the only components of the heterotrimeric GTP binding proteins which harbor known mutations, b/g subunits should be considered possible targets of genetic alterations as suggested by the frequent presence of b3 subunit variants in patients with essential hypertension.