The dynamics of sulfur cycling in wetland peat along an elevational transect at high, intermediate, and low locations (MS, NW, and LB sites, respectively) was investigated in the watershed of McDonalds Branch within the New Jersey Pinelands, by utilizing both soil incubation experiments (with (SO42-)-S-35 as a radiotracer) and stable isotope (delta S-34 and delta O-18) analyses of soil, rain, and streamwater. The results indicate that sulfur cycling can vary greatly among different portions of the watershed and this can have large effects on streamwater chemistry and delta S-34 values over distances as short as 1 km. Laboratory incubations of peat samples collected in July 1993 revealed the co-occurrence of dissimilatory sulfate reduction (DSR; rates ranging from 0.2 to 22.1 nmol/wet g/day) and net generation of sulfate (NaH2PO4 extractable) in the porewater. Generation of sulfate, which was most pronounced at the LB site, may involve oxidation of reduced sulfur in the pear and/or hydrolysis of ester sulfates (ES). Seasonal changes in streamwater SO42-/Cl- molar ratios were similar at LB and NW, being low during the summer and high in the winter, probably a result of higher rates of DSR within the peat during the summer. Consistent with this, higher delta S-34 values of sulfate in streamwater at NW during the summer are attributable to kinetic isotope effects associated with DSR. In contrast to NW, delta S-34 values of streamwater sulfate at LB were consistently lower, fluctuated little throughout the year, and were most negative during the summer (as much as 9 parts per thousand lower than streamwater at NW 1 km upstream). In comparison to NW, SO42-/Cl- ratios were lower in streamwater at LB throughout most of the year except for reversals during the summer, which coincided with the lowest delta S-34 values. In addition, there was a marked difference in the relationship of delta O-18 vs. delta S-34 Of sulfate in LB and NW streamwater, further suggesting that sulfur cycling varies greatly over relatively short distances within this watershed. In order to explain some of these site-specific differences in streamwater SO42-/Cl- ratios, delta S-34, and delta O-18 values, we hypothesize that the ES pool at LB may, by means of hydrolysis or isotopic exchange with streamwater sulfate, serve as an additional source of isotopically light sulfate to streamwater throughout most of the year. During the summer, drier conditions lower the water table at LB and enhance oxidation of reduced sulfur which releases a pulse of even isotopically lighter sulfate to the stream. Copyright (C) 2000 Elsevier Science Ltd.