Dissolution kinetics of arsenic from soils and tailings were studied under simulated gastrointestinal conditions to determine the effects of residence time, pH and soil composition on the bioaccessibility of arsenic. The samples were sieved to four particle size fractions from bulk to <45 μm, and included arsenic minerals, soils and tailings with total arsenic concentrations ranging from 19 to 42000 mg kg(-1). The bioaccessible arsenic concentrations varied from 2.8 to 10000 mg kg(-1), and the highest concentrations were associated with the smallest particle size fractions. Kinetic parameters were determined for each sample extracted under gastric conditions (pH=1.8) followed by intestinal conditions (pH=7.0). Under gastric pH conditions, dissolution appeared to be diffusion-controlled and followed an exponential curve, whereas a logarithmic or linear model was used to describe the mixed dissolution mechanisms observed under intestinal conditions. Nine of the 13 samples tested reached a steady state bioaccessible arsenic concentration within the 5-h physiologically-based extraction test (PBET). However the bioaccessible arsenic concentrations in four tailings samples increased significantly (p=0.034) between the 5-h and the extended 24-h extraction under intestinal conditions. Since arsenic absorption may occur along the entire digestive tract, assessments based on the standard 5-h PBET extraction may not adequately estimate the risks associated with arsenic absorption in such cases. The slow dissolution kinetics associated with secondary arsenic minerals in some tailings samples may require extending the PBET extractions to longer periods, or extrapolating using the proposed kinetic models, to reach steady state concentrations in simulated gastrointestinal fluids.