According to the Rosenblueth-Simeone model, the heart rate (HR) is proportional to the sympathovagal balance. The individual proportionality constant is the intrinsic HR, which can be determined only invasively. The percentage low-frequency spectral HR variability power, relative to the low- plus high-frequency spectral power (%LF) has been raised as a noninvasive alternative. We previously studied young healthy male subjects, in whom gradual autonomic changes were induced by incremental head-up tilt (0-10-20-30-40-45-50-55-60-65-70-75-80 degrees). At each tilt angle we computed HR and %LF. Linear regressions of %LF on HR, characterizing individual autonomic dynamics, confirmed that, within a subject, changes in %LF were proportional to changes in HR. For the current study, we made repeated measurements in 19 subjects after 1 to 8 months. In six subjects, the session 1 and session 2 regression lines differed significantly (t-test, p < 0.05), demonstrating the time dependence of the autonomic dynamics. In such cases, similar HR values on different days are to be associated with different %LF values. We also determined the reproducibility of the supine HR and %LF values. For all 19 subjects, the coefficients of variation were 7 and 22%, respectively: HR reproduces better than %LF. Hence, time-dependent autonomic dynamics contribute systematically to the inferior reproducibility of %LF.