There are over 150,000 contaminated sites in Australia. In most cases, soils are polluted with mixtures of contaminants, including metal/metalloids. However, current human health assessment, particularly in relation to health investigation level guidelines is based on single contaminant. If no site-specific data are available, additive effect is assumed which may overestimate or underestimate the risk. Therefore, a more detailed study on the mixture of metal/metalloids is needed. The emphasis of this study is laid on the exposure risk of arsenic (As), cadmium (Cd), and lead (Pb) when they are ingested as mixtures. Two in-vitro models are adopted in this study. Model 1 is the Unified BARGE (Bioaccessibility Research Group of Europe) method (which is known as UBM) with the aim to measure the soluble fraction of soil contaminants in the human gastrointestinal tract (bioaccessibility) as well as the interaction effects of As, Cd and Pb on respective bioaccessibility. Model 2 is using human liver carcinoma cell line (HepG2) to study the subsequent uptake of contaminants into cells in target organ and possible interaction of As, Cd and Pb on their respective uptake after they are solubilised in the human digestive system. Considering at some contaminated sites polycyclic aromatic hydrocarbons (PAHs) may co-exist with As, Cd and Pb, effects of four selected PAHs (NAP for naphthalene, PHE for phenanthrene, PYR for pyrene, and B[a]P for benzo[a]pyrene) on the bioaccessibility and subsequent uptake of As, Cd and Pb into HepG2 cells are also preliminarily investigated. The focus of this PhD study is on mixtures of As, Cd and Pb.Seven chemically-variant top soils (0-20 cm) were collected in Australia. Soils were characterised for pH, total carbon, total nitrogen, total sulphur, total organic carbon, total inorganic carbon, cation exchange capacity, oxalate-extractable iron, aluminium and manganese, particle size distribution, electrical conductivity etc. A range of concentrations of As, Cd and Pb was spiked into these seven types of soils and aged for up to 12 months. UBM results show the bioaccessibility of As in the gastric phase was not significantly different from those in the intestinal phase whilst a pronounced difference was observed between Cd gastric bioaccessibility and intestinal bioaccessibility as well as between Pb gastric bioaccessibility and intestinal bioaccessibility. Organic carbon, iron oxide and aluminium oxide were key parameters influencing the bioaccessibility of As (gastric and intestinal phases), Cd (intestinal phase) and Pb (intestinal phase).For mixture studies, two exposure scenarios were investigated. Under the first scenario, As, Cd and Pb were spiked to individual soils to mimic independent ageing of mixed 3 contaminants, such as contaminants age at different spots or there has been a long time gap between different contaminants entering the same spot. The interaction between binary mixtures of As, Cd and Pb in simulated human digestive system was studied by mixing soils spiked with A...