Systemic arterial compliance, a major component of aortic input impedance, was determined in 10 patients with congestive heart failure secondary to idiopathic dilated cardiomyopathy and 11 age-matched control subjects found free of detectable cardiovascular disease. Total arterial compliance was determined from high-fidelity ascending aortic pressure and velocity recordings using 1) the traditional monoexponential aortic diastolic pressure decay and 2) the direct solution of the equation, which describes the three-element windkessel model of the arterial system. Resting values for total arterial compliance (x10(-3) cm5/dyn) derived from method 1 were significantly correlated with compliance derived from method 2 (r = 0.89, P less than 0.01). However, method 1 values (control mean 1.15 +/- 0.27, heart failure mean 1.18 +/- 0.54) were consistently and significantly lower (P less than 0.001) than method 2 values (control mean 1.59 +/- 0.50, heart failure mean 1.38 +/- 0.60). Resting total arterial compliance in heart-failure patients was not significantly different from control subjects. Total arterial compliance did not significantly change with exercise in either group despite increases in arterial pressure. However, nitroprusside administration in the heart-failure group increased total arterial compliance both at rest and on exercise compared with the unmedicated state. These different methodological approaches to the estimation of total arterial compliance in humans resulted in significantly different absolute values for compliance, although both methods provided concordant results with respect to the response of arterial compliance to physiological and pharmacological interventions.(ABSTRACT TRUNCATED AT 250 WORDS)