An effective and lossy compression algorithm of multispectral remote sensing images was proposed.On the basis of analyzing characteristic of multispectral images,IHS transform and integer wavelet transform were combined to compress data.IHS transform approach was applied to remove the correlation between spectral bands and integer wavelet transform was applied to remove the correlation of spatial.According to the important degree of the each subband,adaptive threshold values were chosen in the quantification,and fixed bit-plane coding and RLE were individually used to quantify data of every subband and important graph.The experiment result shows that the algorithm has higher compression ratio while PSNR≥33dB.In addition,the algorithm requires small storage and is easy to be realized in hardware.