To characterize the respiratory sinus arrhythmia mechanism (RSA) as a system, in 19 healthy subjects we assessed the effects of continuous and linearly increasing respiratory frequency (RF) from 0.05 to 0.8 Hz at constant tidal volume, in sitting (SIC) and standing (STC) conditions, on: the 90-s time-courses of the central frequency and power of the high-frequency components of RR (CFHFRR, PHFRR) and respiration (CFHFRES, PHFRES), estimated by a time-frequency distribution; the CFHFRES-CFHFRR relation, the time-courses of the difference between CFHFRES and CFHFRR (ΔCFHF) and PHFRR-PHFRES coherence (RSACO) as indexes of RSA input-output coupling (RSAI-OC), and alpha index (square root of PHFRR/PHFRes) as RSA sensitivity (RSAS) measure. The responses of RSA measures to chirped RF in SIC presented three stages with distinctive effects and RF ranges: between 0.09-0.18 Hz, CFHFRR was unchanged, ΔCFHF=-0.10±0.03 Hz, RSACO of 0.77±0.06 and RSAS raised abruptly; between 0.18-0.51 Hz, CFHFRR, ΔCFHF (0.02±0.04 Hz) and RSAS changed proportionally to RF, with correlations of 0.97±0.03, 0.86±0.10 and -0.81±0.12, regression intercepts of 49±47 Hz, -0.87±0.33 Hz, 245±115 ms/l respectively, and RSACO of 0.92±0.03; between 0.51-0.82 Hz, correlations of CFHFRR, ΔCFHF (0.32±0.06 Hz) and RSAS with RF were -0.24±0.64, 0.79±0.08, and -0.36±0.67 respectively, with RSACO=0.78±0.08. STC decreased RSACO (p<0.02), correlations (p<0.01), and intercepts (p<0.03) of CFHFRES-CFHFRR, RF-ΔCFHF and RF-RSAS relations in the 0.18-0.51 Hz range. In SIC and STC, RSAIOC and RSAS measures vary as a function of RF, showing three consecutive stages of some change, proportional change, and no change, in distinctive RF ranges, specifically: in the low RF range, reduced RSAIOC and RSAS overshoot; in the medium one, strong correlations of RF with all RSA measures and high RSACO, although CFHFRR is smaller than RF; and in the high RF range, decreased RSAIOC and RSAS. Baroreflex activation significantly depresses RSAIOC and RSAS measures in the three stages.