A new destriping technique based on wavelet analysis and its application to satellite imagery is presented. The method was tested on a heavily striped Landsat MSS image using Daubechies wavelets of different orders. Qualitative and quantitative analyses were carried out to evaluate the performance for each wavelet, both by visual inspection and by measuring the signal-to-noise ratio of the denoised images. The outcome demonstrates the viability of the wavelet transform as a destriping tool by significantly reducing striping noise. Furthermore, the method proved to defeat some of the problems commonly found with traditional destriping techniques, such as the Gibbs phenomenon, border effects, edge blurring, and the preservation of radiometric level. In addition to its relatively easy implementation, the method is inexpensive in computer time and storage space, as well as suitable for application to problems other than the field of remote sensing.