A test method for the determination of PAH relative bioaccessibility from contaminated soils was implemented and validated for use in risk assessment of soils from contaminated sites meeting performance requirement set for evaluation against regulatory limits. Relative bioaccessibilities of soil benzo(a)pyrene obtained with the test were linearly correlated to relative bioavailability data obtained previously in in vivo investigations with experimental animals, but the relative bioaccessibilities were lower than the relative bioavailabilities. Tests of soil samples from 4 Danish sites contaminated with benzo(a)pyrene and dibenz(a,h)anthracene from different sources and with different ages demonstrated that using the bioaccessible concentrations in risk assessment instead of total concentrations would have produced lower risk classifications of the sites. An urgent requirement is demonstrated for an accepted and valid method for investigating the bioavailability of organic soil contaminants such as PAH and also for soil samples with accepted relative bioavailability data.