Degrading the quality and intelligibility of the speech signals, background noise is a severe problem in communication and other speech related systems. In order to get rid of this problem, it is important to enhance the noisy speech signal mainly through noise reduction. All most all the speech enhancement methods modify the frequency domain spectrum of the noise-corrupted speech to suppress the noise. Although both magnitude and phase spectra together contain the frequency domain information, the traditional speech enhancement procedures either work with magnitude or phase spectrum. This paper presents a speech enhancement method that exploits both magnitude and phase spectra. Experimental studies of the proposed method exhibits better PESQ (Perceptual Estimation of Speech Quality) score than that of other existing methods. We also found better speech quality with our proposed method in several subjective experiments.