Though oxidative stress is recognized as an important pathogenic mechanism of arsenic, and arsenic methylation capacity is suggested to be highly involved in arsenic-related diseases, the association of arsenic methylation capacity with arsenic-induced oxidative stress remains unclear. To explore oxidative stress and its association with arsenic methylation, cross-sectional studies were conducted among 208 high and 59 low arsenic-exposed subjects. Levels of urinary arsenic species [inorganic arsenic (iAs), monomethylated arsenic (MMA) and dimethylated arsenic (DMA)] were determined by hydride generation atomic absorption spectrometry. Proportions of urinary arsenic species, the first methylation ratio (FMR) and the secondary methylation ratio (SMR) were used as indicators for arsenic methylation capacity. Urinary 8-hydroxy-2'-deoxyguanosine (8-OHdG) concentrations were analyzed by enzyme-linked immunosorbent assay kits. Reduced glutathione (GSH) levels and superoxide dismutase (SOD) activity in whole blood were determined to reflect anti-oxidative status. The high arsenic-exposed children and adults were significantly increased in urinary 8-OHdG concentrations but decreased in blood GSH levels compared with the low exposed children and adults. In multiple linear regression models, blood GSH levels and urinary 8-OHdG concentrations of arsenic-exposed children and adults showed strong associations with the levels of urinary arsenic species. Arsenic-exposed subjects in the lower and the upper quartiles of proportions of urinary arsenic species, FMR or SMR were significantly different in urinary 8-OHdG, blood GSH and SOD. The associations of arsenic methylation capacity with 8-OHdG, GSH and SOD were also observed in multivariate regression analyses. These results may provide linkage between arsenic methylation capacity and oxidative stress in humans and suggest that adverse health effects induced by arsenic are related to arsenic methylation through oxidative stress.