Pulmonary hypertension on heart failure (HF) limits exercise capacity and survival probably because of associated right ventricular (RV) failure. This study investigated the mechanisms of RV function adaptation to early pulmonary hypertension in experimental HF. Seven weeks of rapid ventricular pacing in six dogs induced a HF characterized by cardiomegaly and decreased left ventricular ejection fraction. Compared with eight control dogs, pulmonary hypertension was borderline, with a mean pulmonary artery pressure increased to only 23 ± 2 (means ± SE) mmHg. However, the pulmonary vascular impedance spectrum was globally shifted to higher pressures, with an increase in 0 Hz impedance (resistance) to 662 ± 69 vs. 455 ± 41 dynes·cm(-5)·m(2) in controls (P < 0.01) and in characteristic impedance to 183 ± 20 vs. 104 ± 7 dynes·cm(-5)·m(2) in controls (P < 0.01). There was no change in RV end-systolic elastance (Ees), but arterial elastance (Ea) was increased to 1.8 ± 0.3 vs. 0.9 ± 0.1 mmHg/ml in controls so that RV-arterial coupling defined by the Ees-to-Ea ratio (Ees/Ea) was decreased to 0.8 ± 0.1 vs. 1.5 ± 0.1 in controls (P < 0.01). Inhaled nitric oxide, 40 ppm or 5 μg·kg(-1)·min(-1) nitroprusside i.v., did not affect Ees/Ea. Fifty milligrams (i.v.) of milrinone increased Ees/Ea to 1.6 ± 0.2 by an isolated increase in Ees. We conclude that overpacing-induced HF is accompanied by a borderline pulmonary hypertension but profound RV-arterial uncoupling explained by the failure of RV systolic function to adapt combined effects of increased pulmonary arterial resistance and elastance.