Paternity testing is nowadays mostly based on the analysis of DNA short tandem repeats (STR). Number and selection of STR loci differ according to identification kit producers and also particular laboratories. This article refers a revision of formerly excluded paternity. The cause of the mismatch was revealed by the enlargement of STR loci panel and reciprocal evaluation of samples and paternity was on the contrary confirmed. Different possibilities of failures, their consequences and preventions are also discussed.