The factors associated with high concentrations of circulating plasma immunoreactive endothelin in patients with diabetes mellitus are unknown. Plasma and tissue (lung and kidney) immunoreactive endothelin levels were therefore measured by radioimmunoassay in three animal models of diabetes mellitus: dexamethasone-treated rats (2 mg/kg per day for 12 days), streptozotocin-treated rats (100 mg/kg, 4 days before being killed) and rats treated with both dexamethasone and streptozotocin. Plasma concentrations of immunoreactive endothelin in the dexamethasone-treated rats (3·13±0·28 pmol/l, mean ± s.e.m., n = 15) were significantly (P < 0·005) higher than those in controls (1·33±0·18 pmol/l, n = 15), while plasma concentrations of immunoreactive endothelin in streptozotocin-treated rats (n = 8) and rats treated with both dexamethasone and streptozotocin (n= 14) were undetectable (< 0·5 pmol/l). Fast protein liquid chromatographic analysis of the plasma immunoreactive endothelin of dexamethasone-treated rats showed four peaks: one in the void volume, one eluting before endothelin-3, one eluting after endothelin-3 and before endothelin-1 and one eluting in a position identical with that of endothelin-1. Pulmonary concentrations of immunoreactive endothelin in the three groups of rats with diabetes mellitus were lower (P < 0·005) but no significant change was found in renal immunoreactive endothelin. These findings indicate that short-term dexamethasone treatment increases plasma levels of immunoreactive endothelin while streptozotocin treatment decreases them. Thus, multiple factors may influence plasma concentrations of immunoreactive endothelin in diabetes mellitus. Journal of Endocrinology (1991) 130, 123–127