An adaptive wavelet packet transform (AWPT) algorithm is proposed to process synthetic aperture radar (SAR) imagery and remove background clutters from target images. Since target features are more efficiently represented using the wavelet packet bases, higher signal-to-clutter ratios (SCR) can be achieved in the wavelet transform domain. Consequently, clutters can be more effectively separated from the desired target features in the transform domain than in the original SAR domain. The processed results based on the MSTAR data set demonstrate the effectiveness of this algorithm for SAR clutter reduction.