Different animals and biomarkers have been used to measure the relative bioavailability of arsenic (As-RBA) in contaminated soils. However, there is a lack of As-RBA comparison based on different animals (i.e., swine and mouse) and biomarkers [area under blood As concentration curve (AUC) after a single gavaged dose vs steady-state As urinary excretion (SSUE) and As accumulation in liver or kidney after multiple doses via diet]. In this study, As-RBA in 12 As-contaminated soils with known As-RBA via swine blood AUC model were measured by mouse blood AUC, SSUE, and liver and kidney analyses. As-RBA ranges for the four mouse assays were 2.8-61%, 3.6-64%, 3.9-74%, and 3.4-61%. Compared to swine blood AUC assay (7.0-81%), though well correlated (R(2) = 0.83), the mouse blood AUC assay yielded lower values (2.8-61%). Similarly, strong correlations of As-RBA were observed between mouse blood AUC and mouse SSUE (R(2) = 0.86) and between urine, liver, and kidney (R(2) = 0.75-0.89), suggesting As-RBA was congruent among different animals and end points. Different animals and biomarkers had little impact on the outcome of in vivo assays to validate in vitro assays. On the basis of its simplicity, mouse liver or kidney assay following repeated doses of soil-amended diet is recommended for future As-RBA studies.