Fractal image compression is based on the representation of an image by contractive transforms whose xed points are close to the original image. In the encoding of fractal image compression, most of the time is spent on nding the close match between a range block and a large pool of domain blocks. In this thesis, we use the intra-block variances distributions of domain blocks to reduce the searching space. For nding a close match we need only search the domain blocks whose maximal intra-block variance quadrants are the same as that of the range block in nding a close match. The experiment results show that our algorithm can reduce much encoding time with only slight loss of quality.