ABSTRACT: Microarray data are expected to be of significant help in the development of efficient cancer diagnoses and classification platforms. The main problem that needs to be addressed is the selection of a small subset of genes from the thousands of genes in the data that contributes to a cancer disease. This selection process is difficult due to the availability of a small number of samples compared to the huge number of genes, many irrelevant genes, and noisy genes. Therefore, this paper proposes a cyclic method based on genetic algorithms (GA) to select a near-optimal (small) subset of informative genes that is relevant for cancer classification. The performance of the proposed method was evaluated by three benchmark microarray data sets and obtained encouraging results as compared with other experimented methods and previous related works.