Comparative genomic hybridization (CGH) is a relatively new technique whose application is increasing. The method has mostly been employed for detection of chromosome aberrations in cancers, and a large amount of data in this field is accumulating. At the same time, efforts are made to improve the technique in order to increase the sensitivity and the generation of reliable results. Based on experimental data, we have developed a computer algorithm for eliminating some of the interfering effects of unsuppressed repetitive sequences in CGH analysis, and thereby improved our CGH analysis system.