We report time-series data collected over two years for delta18O, delta2H, and Ca, Mg, K, and Cl, concentrations for 10 ponds in, and upflow of, an As-polluted region of southern West Bengal. We compare the compositions of As-polluted groundwaters from wells with the compositions of waters in ponds upflow, and within the range of influence, of the wells. Conservative tracers (delta18O, delta2H, K), and other tracers (Ca, Mg) that are likely conservative in the waters, showthat pondwater and groundwater are distinct and do not overlap in composition. These data show that water from ponds cannot be identified in As-polluted groundwater, so putative DOC in pondwater cannot be mixing into the As-polluted groundwater we have sampled. Separate estimates of the degree of recharge from ponds to groundwater, using calculations based on temporal variations in salt content and isotopic composition in ponds, and salt-balance, show that insignificant amounts of As-polluted groundwater are derived via pond recharge. It follows that pondwater in the study area does not contribute significant mass to arsenic-polluted groundwater and so does not provide organic matterto aquifers in amounts sufficientto drive reduction of iron oxyhydroxides and hence arsenic pollution.