Contamination of food with arsenics is a potential health risk for both humans and animals in many regions of the world, especially in Asia. Arsenics can be accumulated in humans, animals and plants for a longer period and a long-term exposure of humans to arsenics results in severe damage of kidney, lever, heart etc. and many other vascular diseases. Arsenic contamination in human may also lead to development of cancer. In this paper we report our results on data mining approach (an in silico analysis based on searching of the existing genomic databases) for identification and characterization of genes that might be responsible for uptake, accumulation or metabolism of arsenics. For these in silico analyses we have involved the model plant Arabidopsis thaliana in our investigation. By employing a system biology model (a kinetic model) we have studied the molecular mechanisms of these processes in this plant. This model contains equations for uptake, metabolism and sequestration of different types of arsenic; As(V), As(III), MMAA and DMAA. The model was then implemented in the software XPP. The model was also validated against the data existing in the literatures. Based on the results of these in silico studies we have developed some strategies that can be used for reducing arsenic contents in different parts of the plant. Data mining experiments resulted in identification of two candidate genes (ACR2, arsenate reductase 2 and PCS1, phytochelatin synthase 1) that are involved either in uptake, transport or cellular localization of arsenic in A. thaliana. However, our system biology model revealed that by increasing the level of arsenate reductase together with an increased rate of arsenite sequestration in the vacuoles (by involving an arsenite efflux pump MRP1/2), it is possible to reduce the amount of arsenics in the shoots of A. thaliana to 11–12%.