In vitro assays have been developed to determine metal bioaccessibility in contaminated soils; however, their application to Cd is limited. To assess their suitability to determine Cd relative bioavailability (RBA), Cd-RBA in 12 contaminated soils containing 3.00–296 mg kg − 1 Cd were determined using a mouse model and compared with Cd bioaccessibility data based on four assays including the UBM, SBRC, IVG, and PBET. After being administered feed amended with soil or CdCl 2 for 10-day, the Cd concentrations in the mouse liver and/or kidneys were used as biomarkers to estimate Cd-RBA. Cd-RBA was comparable at 34–90% and 40–78% based on mouse liver and kidneys with RSD of 7.10–8.99%, and 37–84% based on mouse liver plus kidneys with lower RSD of 5.8%. Cadmium bioaccessibility in soils varied with assays, with 61–99, 59–103, 54–107, and 35–97% in the gastric phase and 20–56, 38–77, 42–88, and 19–64% in the intestinal phase of the UBM, SBRC, IVG and PBET assays. Based on the combined biomarker of liver plus kidneys, better correlation was observed for PBET (r 2 = 0.61–0.70) than those for IVG, UBM and SBRC assays (0.12–0.52). The monthly Cd intake in children was 0.24–23.9 μg kg − 1 using total Cd concentration in soils, which was reduced by 43% to 0.18–12.3 μg kg − 1 using bioavailable Cd. Our data suggest it is important to consider Cd-RBA to assess risk associated with contaminated soils and the PBET may have potential to predict Cd-RBA in contaminated soils. [ABSTRACT FROM AUTHOR]