Examination of mutations by random amplified polymorphic DNA (RAPD) (also called arbitrarily primed PCR = AP-PCR) in the breast and Wilm's human cancer tissues as well as in estrogen-induced hamster kidney cancer tissues revealed a gain or loss of amplified fragments sizes in breast and Wilm's tumors compared to their respective controls. We also observed changes in the intensity of amplified fragments in tumor DNA compared to control DNA. Most importantly, we found that detection of mutation in the genome of cancer tissue compared to normal tissue by RAPD/AP-PCR depends upon the type of DNA polymerases used. This is the first report showing that polymerase ability to extend a particular locus of the gene is influenced by mutation in the primer binding sites in the competitive environment where several genomic sequences flanked to a random primer are co-amplified by RAPD/AP-PCR. Findings of this study suggest that: i) instead of directly adopting published RAPD protocol it is essential to optimize RAPD reaction conditions by varying magnesium chloride and polymerase concentration with different amounts of DNA templates in order to observe reproducible fingerprints of randomly amplified polymorphic DNA fragments, ii) alterations in the genome as a loss, gain, or decrease and increase in amplification intensity of loci were observed in human tumors as well as in an experimental animal tumor, and iii) before concluding that a mutation is not present when using RAPD/AP-PCR, different polymerases should be used to amplify the genome of the same control or tumor tissues.