Spectral analysis was carried out by Fourier and Gabor transformation of a nonstationary process. It is shown that spectral analysis by Gabor transformation correctly reflects the analyzed signal of varying spectral composition. Fourier transformation is an averaged spectral characteristic. Spectral analysis of anodic current oscillations of copper dissolution in sulfate solution was performed. It was found that current oscillations are a composition of three simple oscillators. The intensity of frequency bands decreases with time. Instantaneous frequencies of each band also decrease in the function of time. Spectral analysis by the Fourier method does not correctly reflect the real changes of analyzed oscillations. Gabor transformation accurately reflects oscillation changes in time. This type of spectral analysis is a useful tool in the analysis of electrochemical oscillations. © 2001 The Electrochemical Society. All rights reserved.