Hypertension has its contribution from atherosclerosis or metabolic factors. It is commonly diagnosed with a sphygmomanometer (blood pressure cuff) which only outputs systolic (Ps) and diastolic (Pd) pressures. Hypertensive patients are then administered oral medication whose effect on arterial compliance (C), an index of vascular stiffness, cannot be accurately determined merely from Ps and Pd values. The underlying blood pressure waveforms are normally obtained clinically through catheterization. Vascular stiffness index (VSI) can then be computed from pulse waveform analysis (PWA). However, there are controversies as to whether a constant single-valued C alone is sufficient for assessing VSI. We investigated the time-varying, pressure-dependent characteristics of arterial compliance (C(P)) with a nonlinear model of the arterial system during normal, induced hypertension and vasodilator-treated conditions. Results show that C(P) differs from constant compliance C and that C(P) can rapidly reflect underlying beat-to-beat variations in arterial compliance and can effectively serve as a reliable index of VSI. It can be used for accurate evaluation of the efficacy of blood pressure altering drugs.