A concrete liner emplaced in 1984 in a shaft at the Waste Isolation Pilot Plant has served as a natural laboratory for observing interactions among concrete, evaporite rocks, and brine. During a routine inspection of the liner in the spring of 1990, discoloration, deposition of secondary salts, wet areas with exposed aggregate grains, softening of paste, surficial spalling, and cracking were observed locally on the concrete surface of the liner. Some construction joints showed apparent leakage of brine from behind the liner, which was nominally 50 cm thick. Seepage brines were nearly saturated relative to CaCl2 and contained lesser amounts of MgCl2 and KCl, and minor NaCl. The liner surface was locally altered to a 1–2 cm friable hygroscopic layer containing little cement paste; concrete cores (7 or 10 cm diameter) through the liner at depths of 248, 254, 255, and 271 m showed similar degrees of alteration at the liner/rock interface. The most profound alteration of concrete was developed in a ˜7 cm zone sub-parallel to and straddling the construction joint cored at a depth of ˜254.5 m. This zone was extensively microfractured, transected aggregate grains, and contained brucite, gypsum, magnesium hydroxychloride hydrate, and locally calcium chloroaluminate instead of the usual phases of hydrated portland cement. Several mechanisms of chemical degradation have been proposed, the most likely being attack by magnesium ions.