The concentration of sulphate in today's oceans-approximately 28,000 mu mol l(-1)-is maintained by a balance between removal by pyrite burial and evaporite deposition and supply by oxidative weathering and the erosion of sulphate minerals from evaporites(1). Oceanic sulphate concentrations were much lower before the rise of atmospheric oxygen about 2.4 Gyr ago(2). The limited spread of delta S-34 values in sedimentary sulphides from 3.85 to 2.5Gyr ago suggests that microbial sulphate reduction, if it played an important role in the Archaean marine sulphur cycle, must have occurred at sulphate concentrations of 200 mu mol l(-1) or less(3). Here we use sulphur isotope systematics of the 2.7Gyr old volcanogenic massive sulphide ore deposits from Kidd Creek, Ontario, to provide constraints on seawater sulphate concentrations independent of biological considerations. By comparing these values with metal and sulphur budgets from modern hydrothermal settings, we estimate that seawater sulphate concentrations 2.7 Gyr ago were roughly 80 mu mol l(-1). At these levels, the residence time of sulphate was on the order of 200,000 years, sufficiently long to make sulphate a conservative compound in the open ocean, but still short enough to suggest that hydrothermal sulphur fluxes were accompanied by a globally significant sink associated with microbial sulphate reduction.