The determination of arsenic in environmental samples suffers interference by chloride which gives rise to the polyatomic interference of (ArCl)-Ar-40-Cl-35 on As-75. Mathematical correction methods based on the ratios (ArCl)-Ar-40-Cl-37/(ArCl)-Ar-40-Cl-35, (ClO)-Cl-35-O-16/(ArCl)-Ar-40-Cl-35 or (ClO)-Cl-37-O-16/(ArCl)-Ar-40-Cl-35, discussed and compared with each other. The method is applied to the determination of arsenic in treated percolate water from a landfill where arsenic containing waste is landfilled. The results obtained when applying the different equations were in good agreement with each other. The results for the untreated percolate water, which could be diluted far enough to eliminate matrix effects and showed negligible interference from chloride, were in good agreement with results obtained with Capillary Zone Electrophoresis. Hydride generation made it possible to determine As(III) and by difference As(V) in both waters. 99.7% of the arsenic present in the percolate water was As(V), which means that during the landfilling, possibly by the infiltration of rainwater, the arsenic is oxidised because the original industrial waste contained 100% As(III).