Reproducibility is always an important property of markers, but even more important with collaborative projects, involving the generation of data by different labs whose results need to be assembled. To obtain reproducible results, the extraction of purified, high quality DNA is a prerequisite for the majority of the marker techniques. For example, degraded and/or unpurified DNA may affect the amplification or restriction of DNA, resulting in unspecific polymorphisms. Even when purified and high molecular weight DNA is used, RAPDs often fail to show reproducible results. This is because RAPD primers are very short (10 bp), which can result in alterations in their annealing behaviour to the template DNA and the resulting band profiles as a result of small deviations in experimental conditions. Therefore, highly standardized experimental procedures are required when RAPD markers are being used. This implies the need for including repeated samples and also the inclusion of reference genotypes which represent bands of known size. Problems with reproducibility in RAPD analysis could be overcome by focusing on mapped markers for which their inheritance has already been verified.