To assess the effect of nicardipine on left ventricular (LV) diastolic function independent of concurrent effects on loading conditions in patients with LV systolic dysfunction due to coronary artery disease, equihypotensive doses of intravenous nitroprusside and nicardipine were administered to 12 patients with congestive heart failure due to previous myocardial infarction (LV ejection fraction less than 0.40). LV micromanometer pressure and simultaneous radionuclide volume were obtained during a baseline period, during nitroprusside infusion, during a second baseline period and during nicardipine infusion. Mean systemic arterial pressure decreased an average of 21 mm Hg with nitroprusside and 19 mm Hg with nicardipine. A greater decrease in LV end-diastolic pressure was observed with nitroprusside (29 +/- 2 to 15 +/- 2 mm Hg, p less than 0.01) than with nicardipine (29 +/- 2 to 25 +/- 3 mm Hg, p less than 0.05). There was a decrease in the time constant of relaxation during nitroprusside but not during nicardipine infusion. There was enough overlap in LV volumes in the baseline and nitroprusside periods to compare diastolic pressure-volume relations over a common range of volumes in 4 patients, and enough overlap in the baseline and nicardipine periods in 11 patients. The relation was shifted downward in 3 of 4 patients taking nitroprusside and in 6 of 11 patients taking nicardipine. The relation between end-diastolic pressure and volume was not shifted with nicardipine.(ABSTRACT TRUNCATED AT 250 WORDS)