We evaluated reflex cardiac responses mediated by carotid baroreceptors in 14 patients with treated congestive heart failure and 14 age-matched healthy subjects. We used a neck chamber to deliver two types of pressure change: 5 s of continuous 50-mmHg suction and an R wave triggered, ramped neck pressure-suction sequence. Reflex latencies (functions of baroreflex arc duration) were comparable in heart failure patients and healthy subjects. However, the average maximum baroreflex slope (gain) was less in heart failure patients than healthy subjects (2.0 vs. 3.5 ms/mmHg, P less than 0.010), the R-R interval response range was smaller (91 vs. 188 ms, P = 0.002), and the resting R-R interval position on stimulus-response relation (operational point) was significantly (13 vs. 40%, P = 0.001) closer to threshold. Stepwise regression analysis suggested that baseline R-R interval variability, used as an index of ongoing vagal-cardiac nerve traffic, and the inverse of antecubital vein plasma norepinephrine level, used as an index of sympathetic nerve activity, contributed significantly to the prediction of abnormal carotid baroreceptor-cardiac reflex responses. Thus our results suggest that in heart failure patients, carotid baroreceptor-cardiac reflex abnormalities are related significantly to ongoing abnormalities of vagal and sympathetic cardiovascular outflow.