1. In vivo the effects of endothelin-1 (ET-1) are limited by its rapid removal from the circulation and possibly by its metabolism by enzymes such as neutral endopeptidase 24.11, deamidase or carboxypeptidase A. Here, using as a model the isolated perfused mesenteric arterial bed of the rat, we have examined the involvements of these enzymatic activities in the vascular responses to ET-1. 2. Samples of Krebs buffer which had been recirculated through the mesenteric arterial bed for 30 min rapidly destroyed the activity of ET-1 as assessed either by bioassay on rings of rat thoracic aorta or by high performance liquid chromatography (h.p.l.c.). For instance, after 15 min incubation with the recirculated-Krebs solution (recirc-K) the contraction induced by 3 x 10(-9) M ET-1 was reduced by more than 90%. Contractions induced by sarafotoxin 6b (3 x 10(-9) M) were similarly suppressed by preincubation with recirc-K whereas those to Arg-vasopressin (3 x 10(-9) M) were unaffected. 3. The degradation of ET-1 by recirc-K was prevented by 1,10-phenanthroline (10(-3) M), abolished by heating the recirc-K solution to 90 degrees C for 15 min, and reduced by EGTA (5 x 10(-3) M) or ET-1(16-21) (10(-5) M). For instance, in the presence of ET-1(16-21) (n = 6) the contraction induced by ET-1 was reduced by only 40% after 15 min incubation with recirc-K buffer. Leupeptin (3 x 10-4 M), dichloroisocoumarin(5 x 10-5 M), phenylmethyl-sulphonyl fluoride (10-3 M), a combination of bacitracin (300 mg ml-1),bestatin (10-5 M), captopril (10-5 M), phosphoramidon (10-4 M) and thiorphan (10-4 M) or Polypep (aproprietary protein digest) did not inhibit the degradation of ET-1 by recirc-K.4. In experiments examining directly the vascular responses of the isolated perfused mesentery of the rat, the addition of cumulative concentrations of ET-1 to the recirculating Krebs solution caused small concentration-dependent increases in perfusion pressure. The inclusion of ET-1(16-2l), ET-1(17-21), or ET-1(18-21) (10-5M) greatly potentiated these responses, but not those to Arg-vasopressin or methoxamine.The effects of 1,10-phenanthroline or EGTA could not be examined in this system because these agents both depressed non-specifically the vasoconstrictor responses of the mesenteric vascular bed.5. Thus, the rat mesentery releases an enzyme that very rapidly destroys ET-1 or the very closely related peptide, sarafotoxin 6b but not Arg-vasopressin. This enzyme is most probably a metallopeptidase because of its sensitivity to inhibition by 1,10-phenanthroline or EGTA. It is particularly interesting that a simple vascular bed such as the mesentery produces such a powerful endothelin metabolising enzyme. It is tempting, therefore, to speculate that the endothelin degrading enzyme active at neutral pH that- we have found is important in the metabolism of ET-1 throughout the vasculature.