1 Conscious. chronically-instrumented, Long Evans rats were given bolus doses of endothelin-1, endothelin-3 (both at 0.01 and 0.1 nmol kg-1), proendothelin-I [1-38] and proendothelin-3 [1-41] (both 0.1 and 1 nmol kg-1) in order to compare their effects on coeliac haemodynamics, because it has been reported that, in conscious dogs, endothelin-I has paradoxical, prolonged hyperaemic vasodilator effects in this vascular bed. Measurements were made also of mesenteric and hindquarters haemodynamics for comparison. In a separate experiment, endothelin-1 (0.1 nmol kg-1) was given before and 20 min after the onset of an infusion of mecamylamine (50-mu-mol kg h-1) to ensure that the responses measured were not confounded by rapid reflex changes in autonomic activity. 2 None of the peptides caused any increases in coeliac flow or any sustained rises in coeliac vascular conductance, although such changes were clear-cut in the hindquarters vascular bed following the higher dose of endothelin-1 and endothelin-3. In animals treated with mecamylamine the regional haemodynamic effects of the higher dose endothelin-I were not different from those in animals with intact baroreflexes. 3 Although the lower dose of both endothelin-I and endothelin-3 caused less marked coeliac than mesenteric vasoconstriction, this difference was not apparent with the higher dose of the peptides, or with proendothelin-I [1-38]. However, proendothelin-3 [1-41] had less marked coeliac and hindquarters vasoconstrictor effects than proendothelin-1 [1-38], in spite of both peptides causing similar changes in mesenteric haemodynamics. These differences could have been due to regional differences in the conversion of proendothelin-I [1-38] and proendothelin-3 [1-41] to endothelin-1 and endothelin-3, respectively. Our findings contradict the proposition that exogenous proendothelin-3[1-41] is not converted into endothelin-3 in vivo. 4 Since, in contrast to endothelin-1 and endothelin-3, proendothelin-1 [1-38] and proendothelin-3 [1-41] caused only small hindquarters vasodilatations, but large vasoconstrictions, it is feasible that the vasodilator responses to exogenous endothelin-1 and endothelin-3 are pharmacological phenomena, and that, in vivo, the production of endothelins from proendothelins exerts widespread vasoconstrictor effects.