Bioaccessibility of hydrophobic organic contaminants (HOCs) from unintentional ingestion of soil is increasingly assessed with in vitro gastrointestinal models incorporating a sorption sink. In this study, the bioaccessibility of DDTs in contaminated soils (n = 11) was determined using "unfed" unified bioaccessibility method (UBM) and fed organic estimation human simulation test (FOREhST) with/without Tenax as an absorbent. By adding Tenax, the bioaccessibility of DDTs determined using UBM was significantly increased from 4.9-30.6% to 31.6-86.0%. In contrast, the bioaccessibility of DDTs determined using FOREhST without/with Tenax were similar with values of 20.0-60.9% vs 31.5-47.6%, implying that the influence of food components on the absorption efficiency of the sink should not be overlooked. Much high fraction of DDTs (bioaccessibility: 11.7-24.8%) remained in FOREhST supernatant after Tenax collection, suggesting that prediction of bioavailability through bioaccessibility obtained by absorbent needs to be treated with caution when bioaccessibility is determined using a "fed state" in vitro method.