Abstract: The assessment and classification of signal quality play an important role in extracting fine features of signals and identifying radiation source targets. In this paper, the research of S-mode response signals of identification friend or foe(IFF)radiation source targets is focused and a two-level quality assessment and classification method is proposed. The assessment and classification process is introduced detailly from two levels: signal quality preliminary assessment and signal quality precise assessment, achieving accurate assessment and classification of the S-mode response signal quality of IFF radiation source targets, which provides basic support for the subsequent extraction of fine features of IFF signals and the recognition of radiation source targets. The experimental verifications are conducted using target data from multiple civil aviation radiation sources of different quality types. The verification results show that the accuracy for quality assessment and classification of the proposed method reaches 95.4 %, and the effectiveness of feature extraction can be significantly improved by selecting the corresponding quality type signals. Furthermore, compared with the basic method and other existing methods, the classification and recognition accuracy of IFF radiation source targets has been improved by more than 15.0 % and 10.0 % respectively.