Administration of fenoldopam mesylate (FM), a dopaminergic agonist, or of cyclic cAMP phosphodiesterase inhibitors (PDE III), for example theophylline and caffeine, induces arteritis in the rat. In this study we characterized the arteritis induced by UK-61,260, an investigational inotropic agent with vasodilatory properties which displays an inhibitory action on cyclic AMP phosphodiesterase, in comparison with lesions induced by FM. The compounds were administered to Sprague-Dawley rats by intravenous infusion over 24 h (FM and UK-61,260), orally or subcutaneously (UK-61,260); the rats were killed and necropsied for pathological examination at various times between 0 h and 28 days post-infusion. Infusion of UK-61,260 at doses of 100, 300 or 400 mg/kg produced arteritis mainly in the mesenteric arteries and occasionally in the renal, pancreatic, gastric and coronary arteries. There were no arterial lesions after infusion of 30 mg/kg, or after administration of 30, 100 or 200 mg/kg per day subcutaneously for 7 days, or after acute administration of 100, 300, 400 or 600 mg/kg orally. Infusion of rats with 72 or 144 mg/kg FM produced arteritis over a wider range of tissues than did UK-61,260. However, the arterial lesions produced by infusion of either drug have the same initial aspect and a similar evolution with time. Immediately after the end of the infusion, minimal necrosis and haemorrhage occurred in the media only, without involvement of the endothelium or the perivascular space. This indicates that the media of the artery is the primary site of injury. The lesions seen 1 and 3 days post-infusion were characterized by severe medial necrosis and haemorrhage with perivascular acute inflammation and appeared macroscopically as haemorrhagic spots on the vessels. On days 7, 14 and 28 post-infusion, no medial necrosis or haemorrhage were present, while perivascular chronic inflammation and moderate smooth muscle hyperplasia were seen. It appeared, therefore, that the lesions underwent repair in 28 days, but footprints of the damage were still present 28 days post-infusion. The similarity between arteritis induced in rats by fenoldopam or by UK-61,260, at doses inducing PDE III inhibition, is consistent with the view that they have a similar pathogenesis. In our view it is probable that these pharmacologically and chemically distinct drugs trigger an increase in intracellular levels of cAMP which in turn triggers vascular damage. The arterial changes observed in the current study after acute administration may explain the increased incidence of polyarteritis nodosa occurring in long term toxicity studies with FM or PDE III inhibitors.