Concentrations of U, Th and Pb, and the isotopic composition of Pb for whole-rock samples of granitoids show: (1) that open-system behavior is nearly universal in the surface and near-surface environment; and (2) that elemental mobility is possible to depths of several hundred meters. Several identified or at least postulated factors that control U and/or Pb mobility include: (1) the mineralogical sites for U and its daughter products; (2) access of groundwater to these sites; (3) the volume of circulating water; and (4) the chemistry of the groundwater. Studies of granitic samples from peralkaline complexes in the Arabian Shield have shown that most samples lost less than 20% of their U during recent exposure to the near-surface environment. Most of the U in these samples appears to be firmly bound in zircons. In contrast, most surface and shallow drill-core samples of the granite of Lankin Dome (Granite Mountains, Wyoming) have lost ??? 70% of their U. Most of the U in these samples is weakly bound in biotite and epidote-family minerals. The granite recovered during the Illinois Deep Drill Hole Project (Stephenson County, Illinois) is mineralogically similar to the granite of Lankin Dome, but this granite lost radiogenic Pb rather than U, probably as a result of exposure to groundwater that had a markedly different chemistry from that in the Granite Mountains. Studies of the Sherman Granite (Wyoming) and the Go??temar Granite (southeastern Sweden) have shown that U and/or Pb mobility is greatest in and near fractured rock. The greater mobility is interpreted to be the result of both a larger water/rock ratio in the fractured rock and exposure to water over an increased surface area (and consequently a greater number of uranium sites). Several types of geochemical and mineralogic data can be used to identify rock-water interaction in granites; however, if rock samples have favorable radiogenic to common Pb ratios, both the amount and approximate timing of U or Pb mobility can be obtained through the use of isotopic studies. Such information can be extremely important in the search for favorable hosts for containment of radioactive waste. Rocks such as the Go??temar Granite have undergone considerable rock-water interaction, most of which occurred ??? 400 Myr. ago and little in recent times. Thus a search for zones that have experienced only a little interaction with water may provide a misleading prediction as to the ability of such zones to shield radioactive wastes from the modern biosphere. From an isotopic point of view, an ideal candidate for evaluation as a host rock for radioactive wastes would have the following characteristics: (1) a high ratio (> 2) of radiogenic to common Pb in order to optimize precision of the results; (2) a simple two-stage geologic history so that results could be interpreted without multiple working hypotheses; and (3) an originally high percentage (> 50%) of labile U so that the results would be highly sensitive to even small amount of rock-water interaction. These characteristics should produce rocks with marked radioactive disequilibrium in surface samples. The disequilibrium should grade to radioactive equilibrium with increasing depth until zones in which water has not circulated are found. Extensive regions of such zones must exist because UThPb systematics of most analyzed granitoids demonstrate closed-system behavior for almost all of their history except for their recent history in the near-surface environment. ?? 1986.