During ODP Leg 168, two of ten boreholes, ODP Holes 1025C and 1026B, were cased through the sediment section, penetrated basaltic crust that is overpressured, and sealed. In 1999 and 2000 the seals were removed, allowing crustal formation fluids to vent and be sampled. The composition of these fluids is compared to those of basal deep-sea pore waters, which have been the basis for estimating geochemical fluxes from low-temperature ridge flank hydrothermal systems. Estimates for the composition of the major ions in formation fluids based on basal pore waters are within 5% of the values measured in borehole fluids. Similar comparisons for minor and trace elements are not as good; some are reactive in the sediment section, resulting in large uncertainties in the pore water extrapolation, while others are influenced by a variety of contaminants, including steel, grease, drilling muds, and basal sediment. Evidence for contamination includes high dissolved and particulate concentrations of several metals (e.g., Fe, Cu, Co, Zn, and Pb) and measurable changes in concentration during the past four years in response to reaction with basal sediment. This new confidence in estimating the primary composition of formation fluids, coupled with advances in thermodynamic and kinetic models, reveals the possibility of anhydrite precipitation in ridge flank hydrothermal systems at temperatures of ~70¿C. Such new insights allow us to address the timing and conditions under which seawater-crustal reactions occur, leading to more accurate models of crustal evolution.