||Hydrothermal vents play important roles in massive sulfide orebody formation, global geochemical cycling, global heat flux, and possibly the origin of life. In spite of advances in our understanding of surficial processes, still little is known about subsurface vent system processes. In particular, the entrainment of seawater into hydrothermal systems and the evolution of fluid prior to being expelled are poorly understood. Seawater and hydrothermal fluid differ in their rare earth element (REE) and Mg concentrations. Therefore, systematic variations in Mg/Ca and REE may trace the entrainment of seawater into the mound and the evolution of hydrothermal fluid. Subsurface samples from an active hydrothermal vent system on a slow-spreading sediment-free ocean ridge system were recovered for the first time when Leg 158 of the Ocean Drilling Program drilled the Trans-Atlantic Geotraverse (TAG) mound in late 1994. REE and Mg concentrations of seventeen anhydrite mineral separates from distinct spatial regimes in the TAG mound were determined by ICP-MS and flame AAS, respectively. The objectives of this study were to trace fluid evolution in the TAG mound and to compare interpretations from REE and Mg/Ca with interpretations from other tracers of fluid evolution. Mg concentrations TAG anhydrites were low and randomly distributed suggesting no systematic spatial variability. Normalizing TAG anhydrite REE patterns to black smoker fluid concentrations revealed two distinct patterns. The first, termed "Group I", showed a flat pattern while the second, termed "Group II", displayed a flat pattern but with a pronounced negative Eu anomaly. Interestingly, two adjacent bands of anhydrite from a TAG-1 sample showed different patterns. Although all anhydrites analyzed in this study showed hydrothermal fluid-like chondrite-normalized REE patterns with positive Eu anomalies and light REE enrichment, samples from different parts of the TAG mound differed in the magnitude of their Eu anomalies and LREE enrichment. For example, anhydrites from beneath white smoker chimneys (the TAG-2 Area) displayed the largest EuN/EuN and LaN/YbN values (consistent with white smoker fluid values). Anhydrites recovered from the TAG-5 Area and from greater than 100 mbsf beneath the Black Smoker Complex (the TAG-1 Area) displayed the lowest EuN/EuN and LaN/YbN values (consistent with black smoker fluid values). Samples from less than 100 mbsf at the TAG-1 Area displayed moderate to low EuN/EuN and LaN/YbN values (consistent with fluids that lie along the evolutionary sequence between endmember black and white smoker fluids). These results suggested that REE patterns and Mg/Ca ratios did not covary and that Mg/Ca ratios were unsuited as tracers of fluid evolution. Furthermore, neither Mg/Ca ratios nor REE could be used to quantify seawater/hydrothermal fluid ratios. However, systematic spatial variations in anhydrite REE patterns appeared to be reliable tracers of the degree of fluid evolution from black smoker to white smoker fluid. Comparison of REE-based interpretations to interpretations of TAG mound circulation and fluid evolution supported by Sr isotopes, fluid inclusion temperatures, and clay-mineral trace-element geochemistry coupled with high variability in REE concentrations between adjacent samples suggest strong temporal and spatial heterogeneity of fluid circulation and composition within the TAG mound. Although this study and others contribute to preliminary models of TAG mound fluid evolution, the heterogeneity of fluid composition clearly indicate the need for a more comprehensive study. It is important that analyses of all tracers be performed on the same sample suite.