In this study, Fast Fourier transform (FFT) and autoregressive (AR) methods were selected for processing the photoplethysmogram (PPG), electrocardiogram (ECG), electroencephalogram (EEG) signals recorded in order to examine the effects of pulsed electromagnetic field (PEMF) at extremely low frequency (ELF) upon the human electrophysiological signal behavior. The parameters in the autoregressive (AR) method were found by using the least squares method. The power spectra of the PPG, ECG, and EEG signals were obtained by using these spectral analysis techniques. These power spectra were then used to compare the applied methods in terms of their frequency resolution and the effects in extraction of the features representing the PPG, ECG, and EEG signals. Some conclusions were drawn concerning the efficiency of the FFT and least squares AR methods as feature extraction methods used for representing the signals under study.