The 3H-3He dating method is applied in a buried-valley aquifer near Dayton, Ohio. The study area is large, not all sampling locations lie along well-defined flow paths, and existing wells with variable screen lengths and diameters are used. Reliable use of the method at this site requires addressing several complications: (1) The flow system is disturbed because of high pumping rates and induced infiltration; (2) tritium contamination is present in several areas of the aquifer; and (3) radiogenic helium concentrations are elevated in a significant number of the wells. The 3H-3He ages are examined for self-consistency by comparing the reconstructed tritium evolution to the annual weighted tritium measured in precipitation; deviations result from dispersion, tritium contamination, and mixing. The 3He-3He ages are next examined for consistency with chlorofluorocarbon ages; the agreement is poor because of degradation of CFCs. Finally, the 3H-3He ages are examined for consistency with the current understanding of local hydrologic processes; the ages are generally supported by hydrogeologic data and the results of groundwater flow modeling coupled with particle-tracking analyses. ¿ 1998 American Geophysical Union