Modulation recognition plays a key role in analysis of the intercepted signals by electronic warfare receivers. A hybrid approach for modulation recognition is introducd. This approach consists of two main classification and decision algorithms. The first is a pattern recognition algorithm for reconstruction of the constellation shape as a stable modulation signature of the intercepted signal. The second is a threshold detection logic algorithm for identification of the modulation type of the intercepted signal. These two algorithms are connected through a feature extraction procedure The complex envelope, which represents both the in phase and quadrature phase components of the received signal after carrier estimation [1,2,11] is utilized as the input of the pattern recognition algorithm. The proposed algorithm shows a robust behavior for identification of single tone digital modulated signals for a relatively low signal to noise ratio (SNR) compared with the ordinary modulation recognition algorithms. Experimental results are shown for various modulation standards including M-ASK, M-PSK, and M-QAM only for M= 8 or 16, received through an AWGN channel and in the presence of carrier recovery errors.