Limited studies comparing arsenic uptake in Bangladeshi rice varieties point to significant genetic and environment variation. The Additive Main Effect and Multiplicative Interaction (AMMI) model was used to analysis genotype x environment interaction (GEI) of arsenic concentration in rice grain and to identify rice genotypes with low grain arsenic concentration for specific or broad growing environments in Bangladesh. Rice grain samples were collected from on-going experiments for stability analysis of Bangladesh Rice Research Institute (BRRI) released rice varieties in wet and dry seasons at 10 BRRI research stations across Bangladesh. The number of varieties was 18 in the wet season and 20 in the dry season. The rice grain samples were collected at maturity from a 5 m2 area. Oven-dried rice grains were digested with a mixture of HNO3 and H2O2 and analyzed for arsenic by ICP- MS. Variety and GEI had significant effects on grain arsenic concentrations in both wet and dry seasons and the relative variability due to environment was greater than that due to varietals differences in both seasons. Environment and GEI accounted for 70-80% and 10-21% of the total variation, respectively while the variety explained only 9% of the total variation. For both seasons, principal component analysis axes (IPCA1 and IPCA2) were highly significant for grain arsenic and IPCA1 captured the major proportion of GEI. The low and medium yielding dry season genotypes had large positive or negative IPCA1 scores indicating high GEI interaction. The grand mean of grain arsenic for dry season genotypes was higher than that of wet season genotypes. All the varieties accumulated higher arsenic in rice grain in both wet and dry seasons at the Shathkira location. In wet season, all strongly photoperiod sensitive varieties had a lower mean grain arsenic concentration than that for less or moderately photoperiod sensitive varieties.