Indirect evidence exists implicating vascular nitric oxide in the pathogenesis of arterial vasodilation in cirrhosis. In the current study, a coincubation assay to estimate the vascular nitric oxide production was developed and the nitric oxide production by arterial segments of cirrhotic and control rats was assessed. In the assay, measurement of reporter monolayer cell-associated cGMP levels allows the influence of nitric oxide released by arterial segments to be determined. RFL-6 cells served as reporter cells. Nitric oxide production was determined in thoracic aorta and mesenteric arteries of 22 control rats, 10 cirrhotic rats without ascites, and 12 cirrhotic rats with ascites. Basal and bradykinin-stimulated (10 moYL) intracellular content of nitric oxidedependent cGMF' was significantly higher in RFL-6 cells coincubated with aortic segments of cirrhotic rats with (21.3 -c 3.6 pm01/105 cells, P < .05 and 44.7 2 7 . 0 pmol/105 cells, P < .025) and without ascites (15.3 5 3.0 pmoY105 cells, P < .05 and 43.2 i 7.6 pmol/105 cells, P < .05) than in those incubated with aortic segments of control rats (9.7 i 1.3 and 19.5 I 2.5 pmol/105 cells). RFLS cells exposed to bradykinin-stimulated mesenteric arterial segments of cirrhotic rats also showed increased cGMP content (ascitic: 2.73 ? 0.31 pmol/105 cells, P < .005; nonascitic: 2.58 ? 0.51 pmoY105 cells, P < .025) compared with cells exposed to control mesenteric arterial segments (1.28 t-0.15 pmol/105 cells). No differences between cirrhotic and control vessels were observed after endothelium denudation. These results indicate that basal and bradykinin-stimulated vascular nitric oxide production is higher in cirrhotic rats with and without ascites than in control rats in and that the endothelial lining is the site where vascular L-arginine nitric oxide pathway activation takes place in experimental cirrhosis. (HEPA-