We examined rapid ventricular cardiac pacing as a means of inducing heart failure in the dog to establish the sequence and nature of physiologic compensation in this preparation. Seven animals paced at 250 beats/min for 3 weeks (VP1 group) showed an increase in cardiac size from 78.5 +/- 9.5(SD) to 105.8 +/- 13.0 cm2, a reduction in mean arterial pressure from 149 +/- 7 to 130 +/- 21 mm Hg, a fall in cardiac index from 196 +/- 57 to 125 +/- 37 ml/kg/min, and an increase in left ventricular filling pressure from 6 +/- 5 to 22 +/- 9 mm Hg and in right atrial pressure from 2 +/- 2 to 5 +/- 3 mm Hg. An additional series of six animals (VP2 group) was paced until a clear biologic end point for heart failure was reached (average 5.3 +/- 1.9 weeks) and they showed similar but more advanced changes compared with the VP1 group. The changes in cardiac size and hemodynamics in the VP1 and VP2 groups were significantly different from those in parallel studies of 10 sham-operated animals. Plasma norepinephrine and renin activity were unchanged in sham-operated animals, whereas in the VP1 group, plasma norepinephrine rose from 338 +/- 118 to 764 +/- 567 pg/ml (p less than .05), but plasma renin activity did not change. In the VP2 group norepinephrine rose from 471 +/- 285 to 999 +/- 425 pg/ml (p less than .025) and plasma renin rose from 2.1 +/- 1.5 to 8.0 +/- 7.1 ng/ml/hr (p less than .05). There was an excellent correlation between plasma norepinephrine and renin activity before the animals were killed in both the VP1 and VP2 groups (r = .88, p less than .001). No change was evident in atrial natriuretic factor content, as determined by bioassay, in sham-operated or VP1 group animals. However, there was a significant reduction in atrial natriuretic activity from the right atrium that was inversely correlated with the level of right atrial pressure in the VP2 group.