Abstract: A study was conducted with four soils samples (0-15 cm) collected from different levels of arsenic (As) contaminated area where arsenic contaminated irrigation water from shallow tube well was used for rice cultivation to determine the maximum adsorption capacity, energy of adsorption and buffering capacity of As. Arsenic was determined by a Perkin-Elmer Aanalyst 100 atomic absorption spectrophotometer equipped with a FIAS-100 flow injection hydride generation system. The test soils showed a large capacity of arsenic adsorption. The application of arsenic progressively increased the equilibrium solution concentration of arsenic. At the lower levels of equilibrium concentration, the adsorption of arsenic linearly increased, but at greater levels of equilibrium solution concentration the rate of As adsorption decreased. Conventional adsorption equations-Langmuir, Freundlich and Temkin equations were used to describe arsenic sorption characteristics of soils. All soils were found to fit well in all the equations (R² = 0.9052 to 0.9974). Highest and lowest adsorption maxima were observed 2000 mg kg^-1 in soil 2 and 3 and 1111 mg kg^-1 in soil 4 respectively. The highest and lowest arsenic buffering capacity were observed 405 in soil 2 and 185 in soil 4 respectively. The highest energy of adsorption obtained for the soil 4, which showed lowest arsenic adsorption maxima and the lowest energy of adsorption obtained with soil 2 followed by soil 3, which showed the highest arsenic adsorption maxima. Arsenic buffering capacity was positively correlated with the maximum adsorption capacity of the soils (r = 0.99). The As adsorption parameters were highly correlated with clay content, FeO and MnO content of the soils and not with the total As content of the soils.