Sulfide sulfur in mid-oceanic ridge hydrothermal vents is derived from leaching of basaltic-sulfide and seawater-derived sulfate that is reduced during high temperature water rock interaction. Conventional sulfur isotope studies, however, are inconclusive about the massbalance between the two sources because S-34/S-32 ratios of vent fluid H2S and chimney sulfide minerals may reflect not only the mixing ratio but also isotope exchange between sulfate and sulfide. Here, we show that high-precision analysis of S-33 can provide a unique constraint because isotope mixing and isotope exchange result in different Delta S-33 (=delta S-33-0.515 delta S-34) values of up to 0.04 parts per thousand even if delta S-34 values are identical. Detection of such small Delta S-33 differences is technically feasible by using the SF6 dual-inlet mass-spectrometry protocol that has been improved to achieve a precision as good as 0.006 parts per thousand (2 alpha). Sulfide minerals (marcasite, pyrite, chalcopyrite, and sphalerite) and vent H2S collected from four active seafloor hydrothermal vent sites, East Pacific Rise (EPR) 9-10 degrees N, 13 degrees N, and 21 degrees S and Mid-Atlantic Ridge (MAR) 37 degrees N yield Delta S-33 values ranging from -0.002 to 0.033 and delta S-34 from -0.5 parts per thousand to 5.3 parts per thousand. The combined delta S-34 and Delta S-33 systematics reveal that 73 to 89% of vent sulfides are derived from leaching from basaltic sulfide and only 11 to 27% from seawater-derived sulfate. Pyrite from EPR 13 degrees N and marcasite from MAR 37 degrees N are in isotope disequilibrium not only in delta S-34 but also in Delta S-33 with respect to associated sphalerite and chalcopyrite, suggesting non-equilibrium sulfur isotope exchange between seawater sulfate and sulfide during pyrite precipitation. Seafloor hydrothermal vent sulfides are characterized by low Delta S-33 values compared with biogenic sulfides, suggesting little or no contribution of sulfide from microbial sulfate reduction into hydrothermal sulfides at sediment-free mid-oceanic ridge systems. We conclude that S-33 is an effective new tracer for interplay among seawater, oceanic crust and microbes in subseafloor hydrothermal sulfur cycles. (c) 2006 Elsevier Inc. All rights reserved.