Models of steroid-induced hypertension in man and rat have been well characterized but the mechanisms by which ACTH and glucocorticoids raise blood pressure are not fully understood. Recently described paracrine (eg endothelial nitric oxide) and humoral (eg PHF) factors may be important in human essential hypertension. These factors were examined in cortisol-induced hypertension in man and ACTH-induced hypertension in the rat respectively. In man, the haemodynamic effects of ACTH can be attributed to the adrenal production of cortisol, but whether the major rodent glucocorticoid corticosterone is responsible for ACTH-induced hypertension in the rat has not been resolved. This question was examined in these studies. In male volunteers, exogenous cortisol raised blood pressure and suppressed endothelium-dependent vasodilatation, by a mechanism which may be nitric oxide synthase dependent. Although dexamethasone and fludrocortisone also raised blood pressure, attenuation of cholinergic vasodilatation was not observed. From these studies, the data suggest that the effect of cortisol on endothelium-dependent vasodilatation is unique to the endogenous hormone and not reproduced by synthetic agonists of GR or MR. Impaired endothelial vasodilator function may contribute to cortisol-induced hypertension in man. In the rat, exogenous corticosterone, administered in doses to achieve circulating concentrations similar to those observed in the experimental model of ACTH excess, reproduced the haemodynamic and some of the metabolic changes which characterize ACTH-induced hypertension. Further, like ACTH-induced hypertension, corticosterone-induced hypertension was prevented by L- but not D-arginine, and this effect was completely prevented by NOLA. It is likely that adrenal corticosterone mediates the hypertensive effects of ACTH excess. Parathyroidectomy had no significant effect on the rise in blood pressure secondary to ACTH excess. It is unlikely that PHF contributes to the model of ACTH-induced hypertension in the rat. The bioassay for the measurement of PHF could not be reproduced in our laboratory, leaving a question mark about the relevance of this putative factor in hypertension research.