Synthetic Aperture Radar (SAR) is a satellite imaging technology. It produces high resolution images and can be used in varied atmospheric conditions. It can be used under all weather conditions. SAR images are generally affected with granular noise termed as speckle noise. Speckle degrades the spatial resolution, contrast of the image, decreases object detectability and causes inconvenience to the SAR image interpreter. Therefore for proper interpretation, it is essential to reduce the speckle noise in SAR images. This paper presents a new speckle noise reduction technique using modelling wa zvelet coefficients. Noiseless and noisy wavelet coefficients are modelled as two-sided generalised Gamma distribution (GΓD) and Gaussian distribution respectively.