The pharmacokinetics and pharmacodynamics of nitroglycerin (GTN) and its dinitrate metabolites, 1,2-glyceryl dinitrate and 1,3-glyceryl dinitrate (1,2-GDN and 1,3-GDN), were examined in a seven-way crossover study in six conscious dogs. The mean apparent clearance after a low GTN i.v. dose (0.025 mg/kg) was 1440 ml/min/kg (S.D. = 630), which decreased to 686 ml/min/kg (S.D. = 317) after a high GTN i.v. dose (0.25 mg/kg), demonstrating dose-dependent pharmacokinetics. Bioavailability of oral GTN (0.25 mg/kg) was very low (0.015 +/- 0.019). The formation of dinitrate metabolites was rapid and extensive after i.v. and oral dosing of GTN. The pharmacokinetics of 1,2-GDN and 1,3-GDN were very similar. The dinitrates exhibited longer half-lives (approximately 45 min) and lower apparent clearances (approximately 16 ml/min/kg) than those found for GTN. The apparent fractional GTN clearances to 1,2-GDN (CLapp,m1) and 1,3-GDN (CLapp,m2) were also determined. The CLapp,m1 decreased significantly (P less than .05) in going from the low GTN dose (1030 +/- 620 ml/min/kg) to the high dose (425 +/- 145 ml/min/kg). In contrast, CLapp,m2 remained unchanged (low dose: 107 +/- 55 ml/min/kg; high dose: 115 +/- 37 ml/min/kg). The duration of systolic blood pressure decrease after low i.v. GTN doses was very short (approximately 15 min) and significantly longer (approximately 90 min) for high i.v. GTN doses. GTN is about 10 to 12 times more efficient than 1,2-GDN and 1,3-GDN in terms of the maximal systolic blood pressure decrease produced by a therapeutic dose. Oral GTN is pharmacodynamically active and this effect can be attributed to the formation of high levels of dinitrate metabolites.