A series of subArctic lakes near Yellowknife, Northwest Territories, Canada, are contaminated with arsenic released from a gold mine. The high environmental arsenic levels afforded an opportunity to identify factors that control methylarsenical production in freshwater sediments. A large variety of methylated arsenicals were observed in most pore water and water column samples from lakes contaminated by aqueous or atmospheric arsenic inputs, albeit at low concentrations relative to inorganic arsenic (generally less than 10% of the total dissolved arsenic concentration). Biologically-mediated methylation/demethylation reactions, therefore, may influence the local arsenic cycle. Water column samples exhibited a consistent methylarsenic composition in which the dominant form was dimethylarsonic acid, followed by monomethylarsinic acid. A much broader range of mono-, di- and trimethylated arsenicals was found in sediment pore water, and the composition of methylarsenic species was highly variable both within and between cores. We provide indirect evidence that many samples contained mono-, di-, and trimethylated arsenic(III)thiols, of the form (CH 3) nAs III(SR) 3−n ( n = 1, 2, 3), the existence of which has been predicted but not previously demonstrated in the environment. The total dissolved methylarsenic concentration in cores consistently exhibited a subsurface maxima between 5 and 15 cm depth, which suggests—in conjunction with the presence of other metabolites in pore water —that methylation may be enhanced by sulfate-reducing bacteria. The role of sulfate-reducers in the production of methylarsenicals observed in sediment pore water is also supported by co-variations with depth between the concentration of methylarsenic species and iron dissolution or other chemical distributions attributed to metabolism by heterotrophic bacteria across different redox zones. Many samples contained appreciable concentrations of arsenicals the specific identity of which are presently unknown, and some of which are ‘hidden’ to conventional hydride-generation analyses. At least one unidentified arsenical found in lake water and sediment pore water appears to have been produced directly or indirectly from atmospheric emissions from the gold mine. The occurrence of methylated arsenicals and hidden arsenic at concentrations that comprise a substantial portion of the total arsenic budget in some aquatic systems suggests that current models of lacustrine arsenic cycling that incorporate only inorganic arsenicals need to be revisited.