1. We have investigated the effects of (i) several guanidines on the activity of the inducible isoform of nitric oxide (NO) synthase (iNOS) in murine cultured macrophages and rat aortic vascular smooth muscle cells (RASM); and (ii) 1-amino-2-hydroxy-guanidine, the most potent inhibitor of iNOS activity discovered, on haemodynamics, multiple organ (liver, renal, and pancreas) dysfunction and iNOS activity in rats with endotoxic shock. 2. The synthesized guanidine analogues caused concentration-dependent inhibitions of the increase in nitrite formation caused by lipopolysaccaride (LPS, 1 microgram ml-1) in J774.2 macrophages and RASM cells with the following rank order of potency: 1-amino-2-hydroxy-guanidine > 1-amino-2-methyl-guanidine > 1-amino-1-methyl-guanidine > 1-amino-1,2-dimethyl-guanidine. Interestingly, 1-amino-2-hydroxy-guanidine (IC50: J774.2, 68 microM; RASM, 114 microM) was more potent in inhibiting nitrite formation caused by LPS than NG-methyl-L-arginine, but less potent than aminoethyl-isothiourea. 3. In the anaesthetized rat, LPS caused a fall in mean arterial blood pressure (MAP) from 115 +/- 4 mmHg (time 0) to 98 +/- 5 mmHg at 2 h (P < 0.05, n = 10) and 69 +/- 5 mmHg at 6 h (P < 0.05, n = 10). The pressor effect of noradrenaline (NA, 1 mg kg-1, i.v.) was also significantly reduced at 1 to 6 h after LPS (vascular hyporeactivity). Treatment of LPS-rats with 1-amino-2-hydroxy-guanidine (10 mg kg-1, i.v. plus 10 mg kg-1 h-1 starting at 2 h after LPS) prevented the delayed hypotension and vascular hyporeactivity seen in LPS-rats. However, 1-amino-2-hydroxy-guanidine had no effect on either MAP or the pressor effect elicited by NA in rats infused with saline rather than LPS. 4. Endotoxaemia for 6 h caused a significant rise in the serum levels of aspartate or alanine aminotransferase (i.e. GOT or GPT) and bilirubin, and hence, liver dysfunction. Treatment of LPS-rats with 1-amino-2-hydroxy-guanidine significantly attenuated the liver dysfunction caused by LPS (P < 0.05, n = 10). Injection of LPS also caused a rapid (almost maximal at 2 h) increase in the serum levels of urea and creatinine, and hence, renal dysfunction. This renal dysfunction was not affected by 1-amino-2-hydroxy-guanidine (P > 0.05; n = 10). Endotoxaemia also caused a dysfunction of pancreas (rise in serum levels of lipase) as well as a metabolic acidosis (falls in PCO2, HCO3 and base excess). Both pancreatic dysfunction and metabolic acidosis were largely attenuated by treatment of LPS-rats with 1-amino-2-hydroxy-guanidine. In rats infused with saline rather than LPS, 1-amino-2-hydroxy-guanidine had no effect on liver, renal or pancreatic function (n = 4). 5. Endotoxaemia for 6 h resulted in a rise in the serum levels of nitrite (11.0 +/- 0.8 microM, P < 0.01, n = 10), which was significantly reduced by 1-amino-2-hydroxy-guanidine (6.5 +/- 0.7 microM, P < 0.05, n = 10). Endotoxaemia for 6 h was also associated with a significant increase in iNOS activity in lung and liver, which was significantly reduced in lung or liver homogenates obtained from LPS-rats treated with 1-amino-2-hydroxy-guanidine. In addition, endotoxaemia for 6 h resulted in a significant increase in myeloperoxidase activity (MPO), an indicator of neutrophil infiltration, in the liver. Treatment of LPS-rats with 1-amino-2-hydroxy-guanidine did not affect the rise in MPO-activity in the liver caused by endotoxin. 6. Thus, 1-amino-2-hydroxy-guanidine is a potent inhibitor of iNOS activity in macrophages or RASM in culture as well as in rats with endotoxic shock. Inhibition of iNOS activity with 1-amino-2-hydroxy-guanidine prevents the delayed circulatory failure and attenuates the dysfunction of liver, and pancreas, as well as the metabolic acidosis caused by endotoxaemia.