1 The haemodynamic dose-response effects of the slow channel blocking agent nicardipine were evaluated in 10 male patients with angiographically confirmed coronary artery disease. At rest, following a similar control saline period, four doses of the drug (log cumulative dosage: 1.25, 2.5, 5.0 and 10.0 mg) were administered by i.v. infusion over a total duration of 40 min; haemodynamic variables were recorded in the 3-5 min following each 5 min infusion. During steady-state exercise the haemodynamic effects of the drug were evaluated by comparison of a control exercise period with observations made at the same workload (200-500 kpm) following the maximum cumulative dose (10 mg).2 Following the four i.v. infusions, the plasma nicardipine level increased log-linearly with the infused dose (r = 0.68). Compared with control measurements at rest after saline, these plasma concentrations (35 + 8 to 141 + 24 ,ug/l) resulted in a linear decrease in systemic blood pressure and vascular resistance with significant increase in cardiac index (maximum ACI + 1.6 1 min1 m-2; P < 0.01), stroke index (maximum ASI + 11 mlIm2; P < 0.01) and in pulmonary artery occluded pressure (maximum APAOP + 2 mm Hg; P < 0.01). There was a significant increase in heart rate; the stroke work index was unchanged. 3 During upright bicycle exercise the reduction in systemic blood pressure was accompanied by an increased exercise cardiac output without change in stroke index. The exercise pulmonary artery occluded pressure was unchanged compared with control observations, the stroke work index fell significantly (P < 0.05). 4 The changes in resting haemodynamic variables following nicardipine reflect the influence of a reduction in left ventricular afterload favourably modifying cardiac performance in ischaemic heart disease. During exercise, with near maximal vasodilatation, the additional reduction in systemic vascular resistance on nicardipine did not further improve exercise haemodynamics. However, nicardipine, over a wide intravenous dose-range, did not result in undue depression of cardiac function; thus it would appear to be haemodynamically safe, even in relatively severe coronary artery disease.