Venous vessels (vena saphena magna) and arteries (left internal mammary artery) from patients treated with long-term ( > 24 h) intravenous infusions of nitroglycerin (CAS 55-63-0, glyceryl trinitrate, GTN) prior to coronary by-pass surgery were studied. Tissue concentrations of GTN and its dinitrate metabolites were determined as well as plasma-concentrations of GTN and glyceryl dinitrates (GDN's) on both the arterial and the venous side. The relaxant effect of GTN on the isolated vein preparations was studied and compared with those from two other groups of patients, one not exposed to GTN and the other exposed to a short-term infusion of GTN. Wide variations were found both in plasma and in tissue concentrations of GTN and GDN's on the venous and the arterial side. However, no extraction across the arterial-venous bed could be observed. Correlations were found between the dose given and the concentrations of drug and metabolites in both venous and arterial plasma as well as between 1,2-GDN in venous plasma and venous tissue, between 1,2-GDN in arterial plasma and arterial tissue and the same applied to 1,3-GDN on the arterial side. No correlations were found between the tissue concentrations of drug or metabolites and the relaxation induced by GTN in vitro. Venous vessels from patients treated with long-term infusions of GTN showed a pronounced decrease in relaxation as compared to vessels from patients not treated with nitroglycerin infusions and a slight decrease as compared to veins from patients treated with a shortterm infusion.