Groundwater pollution with arsenic (As) is occurrence in many countries in Asia such as China, India, Bangladesh, Thailand, Vietnam, and Cambodia. Drinking water is a major pathway that can cause the potential health effects depending on the As species and its concentration, solubility form, magnitude of exposure, and individual factors. Estimation of risk is mainly considered by health risk assessment. Deterministic risk and probabilistic can be used to estimate risk value following the USEPA guide under the considering situation, but Monte Carlo modelling can simulate many cases of output as randomly sampling from probabilistic distribution to make a decision for minimizing risk. Furthermore, the uncertainty of information analysis can be reduced while sensitivity analysis can generate a high effect on the input variable. Both tools can give some different advantages depending on the selection of user and limitation of data resources. Meanwhile, the investigated bioindicator is better able to know the actual level of As in the receptor. The evidence of exposure to other As sources such as the behavior of groundwater consumption, occupation and food consumption should be concerned to reduce the overestimation. Considering exposure scenarios among sources, groundwater contamination, and receptors is a challenge for the relevant organizations. In addition, more attention should be paid to the monitoring and protection plan for vulnerable groups.