An hypothesis of lateral secretion by water of compaction is presented as a possible explanation of the genesis of certain problematical types of ore deposits. These deposits, including sandstone-type V-U deposits of the Colorado Plateau and Mississippi-Valley-type Pb and Zn deposits, have features suggestive of formation by ground water, but proposed explanations of the processes of formation have been widely disputed. It is proposed that water expelled from sediments during compaction may have been the ore-forming fluid. Such water might contain metals in solution prior to burial, as well as metals acquired during diagenesis of the enclosing sediments. Expulsion of the formation fluid must have taken place in huge volume; this volume of fluid, even if highly dilute with regard to ore elements, might have contained sufficient metals to have formed the numerous, widely distributed deposits characteristic of the Colorado Plateau and Mississippi Valley regions. Expulsion of the fluid probably would take place largely through transmissive zones, which could become the loci of major ore concentrations. Numerous smaller deposits may have been formed from smaller-scale expulsion through zones that transmitted less fluid. Flow of formation water during compaction appears to be a more logical source for the widespread small deposits than penetration of consolidated sediments by extrinsic fluids as proposed in previous hypotheses.