The maximum rate of rise of right ventricular pressure (RV dP/dtmax) may change in response to physiological stress and thereby provide an appropriate parameter upon which to base rate adaptive pacing. Initial feasibility testing was carried out in six patients using externally closed loop rate adaptive pacing with a pressure sensing lead (Model 6220) and an investigational VVI pulse generator (Medtronic, Model 2451). During exercise, maximum positive RV dP/dtmax increased from 223 +/- 55 to 405 +/- 181 mmHg.sec.1 (P less than 0.05). Based on these results, rate adaptive pulse generators using maximum positive RV dP/dt were implanted in 12 patients (Medtronic, Model 2503). Exercise treadmill testing in the VVI mode resulted in heart rates ranging from 69 +/- 6 beats/min at rest to 79 +/- 14 beats/min (n = 12; P greater than 0.05). In contrast, VVIR mode pacing rates ranged from 71 +/- 11 beats/min to 115 +/- 24 beats/min (n = 17; P less than 0.05). Holter recording showed heart rates ranging from 51 +/- 6 to 110 +/- 22 beats/min during activities of normal daily living (n = 9; P less than 0.05). Passive postural tilt resulted in rates of 69 +/- 8 beats/min in the supine position increasing to 74 +/- 14 beats/min with 60 degrees upright tilt (n = 16; P greater than 0.05). With up to 5-year follow-up data, there have been no late failures of pacing but one lead showed insulation failure with over- and undersensing after 4.5 years. A number of deficiencies were identified in the prototypes leading to modifications of a subsequent generation of rate responsive pacemaker based on RV dP/dtmax. These initial data demonstrate that rate adaptive pacing based on RV dP/dtmax responds in a physiological manner. This rate responsive system is of particular interest as it is based on a beat-to-beat parameter of cardiac mechanical function.