To study the differentiation of adrenergic (epinephrine-synthesizing) neurons in brain, the initial appearance and ontogeny of phenylethanolamine N-methyltransferase (PNMT), a specific marker of the adrenergic phenotype, were studied with immunocytochemistry and catalytic assay. The appearance of immunoreactivity to dopamine β-hydroxylase (DBH-IR), an enzyme common to the noradrenergic and adrenergic phenotypes, was also studied. DBH-IR was initially observed on embryonic Day 13 (E13) in cells located on the ventrolateral floor and wall of the rhombencephalon. A day later (E14), PNMT-IR cells and PNMT catalytic activity were observed in the rhombencephalon suggesting that, as in the adrenal gland, noradrenergic expression precedes adrenergic expression. The PNMT-IR cells were presumed to be precursors of C1 neurons since they were located in the ventrolateral medulla oblongata. Cells located in the wall of the medulla which appeared to be migrating ventrally to the C1 group also contained PNMT-IR. On E15, cells which had PNMT-IR processes coursing through the germinal zone were observed dorsally near the fourth ventricle. Although the location of the C1 cell group was apparent when PNMT was initially expressed, the dorsal C2 and C3 adrenergic cell groups were not evident until late in gestation on E19. Even in the term embryo there appeared to be PNMT-IR cells which had not yet reached their final destination. On E14 and E15, PNMT-IR cells were also observed on the floor of the pons just rostral to the pontine flexure. However, these were not observed in older embryos, suggesting that transient expression of PNMT occurs in brain, as well as in the periphery. To determine whether glucocorticoids regulate brain PNMT, we examined the effects of altered glucocorticoid levels. In contrast to PNMT in the sympathetic nervous system, PNMT activity in medulla oblongata was not affected in neonates or adults by the decrease in glucocorticoids following adrenalectomy or hypophysectomy. Conversely, elevation of glucocorticoids by hormonal treatment did not alter PNMT in neonates. Notably, however, treatment of pregnant rats with dexamethasone on E18–E21, but not earlier, increased PNMT activity in the fetal brain stem. These observations suggest that PNMT expression and development is regulated by different factors in cells derived from neural crest and tube. PNMT is expressed earlier in brain than in adrenal and sympathetic ganglia. Further, the development of PNMT in the periphery, but not in the brain, is dependent on maintenance of physiological levels of glucocorticoids. Our studies also elucidate three similarities in central and peripheral PNMT development: (1) in both adrenal and brain, adrenergic expression follows noradrenergic expression, (2) in both brain and periphery, PNMT is transiently expressed in some cells, and (3) glucocorticoid treatment of pregnant rats increases PNMT in the embryonic brain and sympathetic nervous system. Finally, we have observed a critical period during which glucocorticoid treatment increases PNMT activity in the brain.