In 18 patients with obstructive sleep-apnea syndrome (OSAS), time-variable spectral analysis was employed to study the course of heart-rate variability (HRV) and blood-pressure variability (BPV), before and during apnea episodes. Time intervals of 6 minutes' length were detected and analyzed from nocturnal sleep recordings. In addition, a software package was developed in MATLAB that calculates time-dependent output spectra by applying short-time Fourier transformation (STFT). Various spectral parameters from HRV and SBV were extracted by means of integration and scaling. In the transition from normal respiration to an apnea episode, significant parameters of HRV and systolic blood pressure variability (SBPV) are increased, whereby alterations were able to be detected even before apnea onset. The time-variant analysis of HRV and SBPV in the frequency range appears to be a suitable method for description of the relationship between the development of apnea and alterations in autonomous tone in patients with OSAS.