Ion-exchange batch experiments were run on Cretaceous (Magothy aquifer) clay cores from a nearshore borehole and an inland borehole on Long Island, NY, to determine the origin of high SO42- concentrations in ground water. Desorption batch tests indicate that the amounts of SO 42- released from the core samples are much greater (980-4700 ??g/g of sediment) than the concentrations in ground-water samples. The locally high SO42- concentrations in pore water extracted from cores are consistent with the overall increase in SO 42- concentrations in ground water along Magothy flow paths. Results of the sorption batch tests indicate that SO42- sorption onto clay is small but significant (40-120 ??g/g of sediment) in the low-pH (<5) pore water of clays, and a significant part of the SO42- in Magothy pore water may result from the oxidation of FeS2 by dissolved Fe(III). The acidic conditions that result from FeS2 oxidation in acidic pore water should result in greater sorption of SO42- and other anions onto protonated surfaces than in neutral-pH pore water. Comparison of the amounts of Cl- released from a clay core sample in desorption batch tests (4 ??g/g of sediment) with the amounts of Cl- sorbed to the same clay in sorption tests (3.7-5 ??g/g) indicates that the high concentrations of Cl- in pore water did not originate from connate seawater but were desorbed from sediment that was previously in contact with seawater. Furthermore, a hypothetical seawater transgression in the past is consistent with the observed pattern of sorbed cation complexes in the Magothy cores and could be a significant source of high SO42- concentrations in Magothy ground water.