Random amplified polymorphic DNA (RAPD) analysis appears to offer a cost- and time-effective alternative to restriction fragment-length polymorphism (RFLP) analysis. However, concerns about the ability to compare RAPD results from one laboratory to another have not been addressed effectively. DNA fragments that were amplified by five primers and shown to be reproducibly polymorphic between two oat cultivars (within the Ottawa laboratory) were tested in six other laboratories in North America. Four of the six participants amplified very few or no fragments using the Ottawa protocol. These same participants were able to generate a considerable number of amplified fragments by using their own protocols. The reproducibility of results among laboratories was affected by two factors. First, different laboratories amplified different size ranges of DNA fragments, and, consequently, small and large polymorphic fragments were not always reproduced. Second, although reproducible results were obtained with four of the primers, reproducible results were not obtained with the fifth primer, using the same reaction conditions. It is suggested that if the overall temperature profiles (especially the annealing temperature) inside the tubes are identical among the laboratories, then RAPD fragments are likely to be reproducible.