1. Peroxynitrite, a potent cytotoxic oxidant formed by the reaction of nitric oxide with superoxide anion, and hydroxyl radical, formed in the iron-catalysed Fenton reaction, are important mediators of reperfusion injury. In in vitro studies, DNA single strand breakage, triggered by peroxynitrite or by hydroxyl radical, activates the nuclear enzyme poly (ADP-ribose) synthetase (PARS), with consequent cytotoxic effects. Using 3-aminobenzamide, an inhibitor of PARS, we investigated the role of PARS in the pathogenesis of splanchnic artery occlusion shock. 2. Splanchnic artery occlusion and reperfusion shock (SAO/R) was induced in rats by clamping both the superior mesenteric artery and the coeliac trunk for 45 min, followed by release of the clamp (reperfusion). At 60 min after reperfusion, animals were killed for histological examination and biochemical studies. 3. SAO/R rats developed a significant fall in mean arterial blood pressure, significant increase of tissue myeloperoxidase activity and marked histological injury to the distal ileum. SAO/R was also associated with a significant mortality (0% survival at 2 h after reperfusion). 4. There was a marked increase in the oxidation of dihydrorhodamine 123 to rhodamine (a marker of peroxynitrite-induced oxidative processes) in the plasma of the SAO/R rats, starting early after reperfusion, but not during ischaemia alone. Immunohistochemical examination demonstrated a marked increase in the immunoreactivity to nitrotyrosine, a specific 'footprint' of peroxynitrite, in the necrotic ileum in shocked rats, as measured at 60 min after the start of reperfusion. 5. In addition, in ex vivo studies in aortic rings from shocked rats, we found reduced contractions to noradrenaline and reduced responsiveness to a relaxant effect to acetylcholine (vascular hyporeactivity and endothelial dysfunction, respectively). 6. In a separate set of studies, using a 4000 Dalton fluorescent dextran tracer, we investigated the changes in epithelial permeability associated with SAO/R. Ten minutes of reperfusion, after 30 min of splanchnic artery ischaemia, resulted in a marked increase in epithelial permeability. 7. There was a significant increase in PARS activity in the intestinal epithelial cells, as measured 10 min after reperfusion ex vivo. 3-Aminobenzamide, a pharmacological inhibitor of PARS (applied at 10 mg kg(-1), i.v., 5 min before reperfusion, followed by an infusion of 10 mg kg(-1) h(-1)), significantly reduced ischaemia/reperfusion injury in the bowel, as evaluated by histological examination. Also it significantly improved mean arterial blood pressure, improved contractile responsiveness to noradrenaline, enhanced the endothelium-dependent relaxations and reduced the reperfusion-induced increase in epithelial permeability. 8. 3-Aminobenzamide also prevented the infiltration of neutrophils into the reperfused intestine, as evidenced by reduced myeloperoxidase activity. It improved the histological status of the reperfused tissues, reduced the production of peroxynitrite in the late phase of reperfusion and improved survival. 9. In conclusion, our study demonstrates that the PARS inhibitor 3-aminobenzamide exerts multiple protective effects in splanchnic artery occlusion/reperfusion shock. We suggest that peroxynitrite and/or hydroxyl radical, produced during the reperfusion phase, trigger DNA strand breakage, PARS activation and subsequent cellular dysfunction. The vascular endothelium is likely to represent an important cellular site of protection by 3-aminobenzamide in SAO shock.