Arsenic contamination in soil has become a concern worldwide. The studies of arsenic forms/species and bioavailability are the foundation for pollution recognition, risk assessment and remediation criteria setting. Methods currently employed for arsenic fractionation and speciation include chemical selective extraction, instrumental detection or resin/cartridges separation after solvent extraction and synchrotron radiation XANES direct detection. The combination of the above various methods have played a core role in the study of arsenic bioavailability. Also, various methods have been used for measurement of arsenic bioavailability in soils. Chemical extraction is cost effective and the most convenient, but limited in not really reflecting the arsenic bioavailability from soils to animals or humans, and usually is used as reference for recognition of arsenic environmental hazards; Indicator plant method rely on successful selection of sensitive plants in order to give clue to the potential hazards of arsenic to environment and health; The population and structure as well as genetic changes of microorganisms exposed to arsenic in soils are direct in interpretation of arsenic toxicity, and are in priority in future application for risk assessment of arsenic contamination. In vitro gastrointestinal method is the most efficient in simulating the state and condition of animal and/or human gastrointestinal system, and is vital in application for risk assessment of arsenic contamination. However, correlation between in vitro and in vivo results is still needed for its accurate interpretation for arsenic risks. Rabbit, young swine and monkey have been commonly used in animal model (in vivo) for measurement of arsenic availability in soils via detection arsenic contents in their urines. However, it is not yet clear which is the most suitable for the reflection of arsenic risks to humans due to limited studies on such kind of studies with relation to human arsenic exposures.