Industrial electrical interference in seismic data has the characteristics of single-frequency, which can be expressed with sine or cosine function. Therefore, this paper presented a new method of automatic identification and self-adaptive suppression of the single-frequency interference based on normalized cross correlation coefficient. We combined the sine function and cosine function to depict the single-frequency interference within the deep time-window, which can approximate the single-frequency interference that may be contained within the practical seismic data, and then minimized the square difference target function between the practical seismic data and the predicted single frequency interference, and finally computed the amplitude, frequency and phase of single-frequency interference. We carried out the normalized cross correlation computation between the predicted single-frequency interference and the practical seismic data, the larger the normalized cross correlation coefficient is, the more possibility the seismic data contains the single-frequency interference. On the contrary, this seismic trace does not contain single frequency interference or contains single frequency interference with week amplitude that can be ignored. Consequently, we presented a threshold value which is used to select the seismic trace containing single-frequency interference, which is subtracted from the original seismic data. With above processing stages, we finished the automatic identification and selt-adaptive suppression of single-frequency interference from the seismic data. The practical seismic data verifies the effectiveness and accuracy of the proposed method, which is better than conventional band pass filter square energy identifying method and the industrial software over the identification accuracy and suppression effect. In a summary, the proposed method can provide an effective method for the practical seismic data amplitude-preserved processing.