Known methods of testing for interference in human genetic maps are reviewed. Two aspects of interference are distinguished, numerical interference referring to the numbers of crossovers occurring and positional interference referring to the positions of crossovers. Tests for interference generally address only one or the other of these two aspects. Published investigations of interference presume fully informative marker maps, while in reality genetic markers are of limited heterozygosity, which weakens some of the tests for interference. The conclusion is drawn that rigorous tests for interference have not yet been carried out for human data.