The Pasco Basin in southeastern Washington State provides a unique hydrogeologic setting for evaluating the chloride mass balance technique for estimating recharge. This basin was affected by late Pleistocene catastrophic floods when glacial dams in western Montana and northern Idaho were breached. It is estimated that multiple Missoula floods occurred between ~13,000 and 15,000 years B.P. and reached a high water elevation of ~350 m. These floods removed accumulated chloride from the sediment profile, effectively resetting the chloride mass balance clock at the beginning of the Holocene. The rate of chloride accumulation qCl in the sediments was determined by two methods and compared. The first method measured qCl by dividing the calculated natural fallout of 36Cl by a measured ratio of 36Cl/Cl in the pore water, while the second method used the total mass of chloride in the profile divided by the length of time that atmospheric chloride had accumulated since the last flood. Although the two methods are based on different approaches, they showed close agreement. In laboratory studies the sediment to water ratio for chloride extraction was sensitive to the grain size of the sediments; low extraction ratios in silt loam sediments led to significant underestimation of pore water chloride concentration. Br/Cl ratios were useful for distinguishing nonatmospheric (e.g., rock) sources of chloride. Field studies showed little spatial variability in estimated recharge at a given site within the basin but showed significant topographic control on recharge rates in this semiarid environment. An extension of the conventional chloride mass balance model was used to evaluate chloride profiles under transient, time-varying annual precipitation conditions. This model was inverted to determine the paleorecharge history for a given soil chloride profile, and the parameters of the root extraction model required to estimate paleoprecipitation. ¿ American Geophysical Union 1996