To determine the constancy of hemodynamic and antianginal effects of the constant infusion of intravenous nitroglycerin (NTG) and their relationship to infusion rate and plasma NTG concentration, we administered maximal tolerated doses of intravenous NTG (range 10 to 120 ,g/min, mean = 52 + 33 ,ug/min) and placebo to 10 patients with chronic stable angina for 25 hr each in a randomized, double-blind fashion. Sublingual NTG (0.4 mg) was given at 24.5 hr of infusion as a positive control. Bicycle exercise time (NIH protocol), blood pressure, heart rate, exercise ST response, and venous plasma NTG were determined before and at 1, 4, 8, 24, and with widely varying slopes and intercepts. Intravenous NTG produced a sustained reduction in blood pressure and a rise in heart rate at rest, and a reduction in blood pressure during submaximal exercise at as late as 24 hr, associated with reduced submaximal ST segment abnormality. In contrast, exercise tolerance to onset of angina showed a marked initial increase on intravenous NTG but fell progressively and did not differ from that with placebo at 24 hr. Increased exercise tolerance was associated with an increase in maximal heart rate and double product (heart rate X blood pressure), suggesting that direct coronary vasodilation and/or reduced left ventricular volume were the principal determinants of increased exercise tolerance. The rates of fall in exercise tolerance over time varied widely among patients. Sublingual NTG produced a marked increase in exercise tolerance after a 24.5 hr placebo infusion, but not after intravenous NTG, despite similar exercise tolerance on intravenous NTG and placebo at 24 hr. The plasma NTG concentrations achieved with intravenous NTG were at least twice those reported for the peak sublingual NTG effect and up to 50 times those reported for 5 mg/24 hr release NTG patches. Thus, constant NTG infusion can result in constant high plasma NTG, but the initial marked increase in exercise tolerance diminishes progressively over 24 hr, as previously observed with NTG patches, consistent with development of tolerance. After prolonged high plasma NTG concentrations, the effect of sublingual NTG on exercise tolerance can be abolished. We conclude that rapid attenuation of antianginal effects during exercise is an inherent result of the continuous administration of NTG. It remains uncertain whether similar tolerance limits the efficacy of intravenous NTG in patients with cardiac ischemia at rest. Circulation 77, No. 6, 1376-1384, 1988 have shown marked attenuation of NTG patch effect over 24 hr, and little or no effect during long-term