The evaluation of speech interference effect refers to the technology of analyzing the disturbed speech received by the communication system to determine speech interference effect level. Accurately evaluating the interference effect is of great significance to the development of communication countermeasure equipment, the assessment of the situation of electronic countermeasures and understanding of communication quality. As for ultra-short wave communication jamming system, a method is proposed based on Mel-Frequency Cepstral Coefficients(MFCC) features, wavelet statistical features and perceptual features, combined with the least squares, the Back Propagation(BP) neural network and Support Vector Regression(SVR) fitting regression model, the correlation coefficient between the predicted value and subjective evaluation value is above 0.9, which guarantees the practicability of the evaluation system. Secondly, the non-reference evaluation method is studied based on deep learning, and the measured data is adopted to verify the effectiveness of this method. The accuracy rate is 87%, higher than that of the multi-measure fusion evaluation method.