The hemodynamic effects of pacing in patients with congestive heart failure (CHF) remain controversial. Early studies reported that pacing from the right ventricular (RV) apex improved acute hemodynamic parameters in patients with left ventricular systolic dysfunction, but these findings were not confirmed in subsequent controlled studies. More recently, it has been proposed that pacing from the RV side of the ventricular septum improves hemodynamic function compared with intrinsic conduction or apical pacing. Either dual-chamber or ventricular pacing have been evaluated, again with inconsistent findings. To assess the effects of pacing site and mode on acute hemodynamic function, we evaluated 21 subjects with CHF and intrinsic conduction disease. Hemodynamics were compared in AAI, VVI, and DDD modes with pacing from the RV apex or high septum. The pacing rate was constant in each patient and the order of testing was randomized. In the absence of ventricular pacing (AAI mode), the mean systemic arterial pressure was 85 +/- 11 mm Hg, the right atrial pressure was 11 +/- 4 mm Hg, the pulmonary capillary wedge pressure was 18 +/- 8 mm Hg and the cardiac index was 2.4 +/- 0.7 L/min/m(2). Compared with AAI pacing, there were no improvements in any hemodynamic parameter with DDD pacing from either RV site. Hemodynamic function worsened with VVI pacing from both RV sites. Subgroup analyses of patients with dilated cardiomyopathy, with prolonged PR interval, or with significant mitral regurgitation also failed to demonstrate an improvement with pacing. We conclude that pacing mode but not RV pacing site affects acute hemodynamic function. Pacing in the DDD mode prevents the deleterious effects of VVI pacing in this patient population.