As an important human exposure pathway of contaminants, soil ingestion is of increasing concern for assessing health risk from polycyclic aromatic hydrocarbons (PAHs) in soils. A wide range of total PAH concentrations ranging from 0.112 microg g(-1) to 27.8 microg g(-1) in soils collected from different public sites, including gas stations, roadsides, bus stops, a kindergarten, primary and middle schools, a university and residential area, was detected. In general, total PAHs concentrations in soils from traffic areas were significantly higher than that from the other sites, indicating a dominant contribution from motor vehicles. Physiologically based in vitro tests were used to evaluate the oral bioaccessibility of PAHs in surface soil under different land uses in Beijing regarding both gastric and small intestinal conditions. It was found that the oral bioaccessibility of total PAHs in small intestinal condition, ranging from 9.2% to 60.5% of total PAHs in soil, was significantly higher than gastric condition, ranging from 3.9% to 54.9%. The bioaccessibility of individual PAHs in soils generally decreased with the increasing ring number of PAHs in both gastric and small intestinal conditions. However, the ratio of bioaccessibility of individual PAHs in gastric condition to that in small intestinal condition, generally increased with increasing ring number, indicating the relatively pronounced effect of bile extract on improving bioaccessibility of PAHs with relatively high ring numbers characterized by their high K(ow) values. The observation that bile extract at a level higher than critical micelle concentration could reduce the surface tension of digestive juice substantially, which may cause PAHs to be available for intestinal absorption, calls for more careful establishment of reliable soil criteria for PAHs, especially concerning the health of children who may ingest a considerable amount of PAH-contaminated soil via outdoor hand-mouth activities.