Published information on the reproducibility of human baroreflex responses in the absence of interventions is limited. Therefore, we analyzed retrospectively vagally mediated carotid baroreceptor-cardiac reflex responses of 34 healthy young adult volunteers whom we studied twice, 7-10 days apart (all 34 subjects) or 10 wk apart (8 subjects). We delivered a sequence of neck pressure changes during held expiration: A computer-driven bellows initially raised pressure to approximately 40 mmHg for five heart beats, and then reduced pressure in a stepwise series of R-wave-triggered 15-mmHg decrements to about -65 mmHg. R-R interval changes were plotted as functions of the carotid distending (systolic less neck) pressure occurring within each interval. Each experimental session yielded one stimulus-response relation, which comprised the average of seven separate trials. Six measures were derived from these relations: minimum, maximum, and range of R-R intervals; maximum slope; and operational point [(R-R interval shortening/R-R interval range) x 100%]. Linear regression correlation coefficients for measurements made on two occasions were all highly significant (range: 0.64-0.99). Our results indicate that human vagally mediated carotid baroreceptor-cardiac reflex responses, studied serially under exacting experimental conditions, are highly reproducible.