The goal of data compression is to find shorter representations for any given data. In a data storage application, this is done in order to save storage space on an auxiliary device or, in the case of a communication scenario, to increase channel throughput. Because remotely sensed data require tremendous amounts of transmission and storage space, it is essential to find good algorithms that utilize the spatial and spectral characteristics of these data to compress them. A new technique is presented that uses a spectral and spatial correlation to create orderly data for the compression of multispectral remote sensing data, such as those acquired by the Landsat Thematic Mapper (TM) sensor system. The method described simply compresses one of the bands using the standard Joint Photographic Expert Group (JPEG) compression, and then orders the next band’s data with respect to the previous sorting permutation. Then, the move-to-front coding technique is used to lower the source entropy before actually encoding the data. Owing to the correlation between visible bands of TM images, it was observed that this method yields tremendous gain on these bands (on an average 0.3 to 0.5 bits/pixel compared with lossless JPEG) and can be successfully used for multispectral images where the spectral distances between bands are close.