The article presents the application of conditional averaging of stochastic signals to determination of correlation interval. For chosen models of signals the results of theoretical analysis are compared with results of experiments. The paper is divided into five sections. The first is a short introduction to the subject of the paper. Section 2 presents the definition and some examples of correlation intervals for typical form of autocorrelation functions (Fig. 1, Tab.1). Section 3 describes the use of conditional mean value to determination of correlation interval (Eq. 9) and statistical errors of estimation for this method (Eq. 10, Eq. 13). The results of experiments for random signals with Gaussian probability distribution and two typical form of autocorrelation function (Fig. 4) are given in Section 4. Section 5 summarizes the results and presents final remarks. The authors conclude that the method described in this paper may be applied to determination of correlation interval of stochastic signals, particularly for signals with non-oscillative form of autocorrelation function.