Baroreflex sensitivity (BRS) has been proposed as a diagnostic parameter for neurological disorders and as a survival-prognosis parameter in diabetic and cardiac patients. Therefore reference values and the reproducibility of BRS were assessed, taking into account the possible influence of age, gender, test conditions and some analysis variants. Healthy subjects ( n = 191) were randomly selected from the 50–75-year-old general population (the Hoorn Study). Variations in blood pressure and heart rate were recorded non-invasively during three breathing modes: spontaneous (3 min), slow metronome (1 min; 6 breaths/min = 0.1 Hz) and fast metronome (1 min; 15 breaths/min = 0.25 Hz), all in a supine position. From these recordings, BRS was assessed as the transfer gain between changes in blood pressure and heart period, and as the α coefficient. BRS values ranged from 5.0 to 8.9 ms·mmHg -1 . Slow metronome breathing resulted in higher BRS values than fast breathing, while during spontaneous breathing BRS in the low-frequency band was lower than that in the high-frequency band (respiratory origin). BRS values decreased with lower coherence criteria. BRS-α was significantly higher than BRS-gain. While regression analysis showed no gender differences, BRS decreased with age. Therefore age-specific reference values were calculated. The reproducibility of BRS values was in general moderate, with reliability coefficients ranging from 43 to 81% and coefficients of variation ranging from 34 to 59%. In conclusion, this study shows age, breathing mode, frequency and coherence threshold to affect measures of BRS. Therefore these factors should be considered in clinical studies; appropriate reference values are given.