The genesis of granitic igneous pegmatites is here considered in terms of a model conceived from results of field and laboratory studies and subsequently tested by means of experimental investigations. This model emphasizes the roles of water (and/or other relatively volatile substances), both as a dissolved constituent in granitic magmas and as the dominant constituent of a separate fluid phase that is in the supercritical state under most conditions of pegmatite formation. Pegmatite magma, as distinguished by a content of dissolved water that is high relative to the limit of solubility under existing confining pressure, can be formed either through partial melting of crustal materials or as rest-liquid in a cooling igneous body yielding dominantly anhydrous crystalline phases. Such granitic magma can be expected to consolidate according to the following three-fold sequence: 1. Crystallization from hydrous silicate melt, yielding anhydrous solid phases with or without OH-bearing phases. The product is characterized by normal phaneritic textures that generafly are coarse grained. It has been termed pegmatite in some occurrences, and granite in others. 2. Crystallization concomitantly from silicate melt and from a coexisting exsolved aqueous fluid of considerably lower viscosity, yielding giant-textured pegmatite along with much finer-grained, even aplitic, mineral aggregates. Segregation of these products can vary enormously in scale and degree. Partitioning of constituents between melt and aqueous fluid, rapid diffusion of constituents through the aqueous phase, and gravitational rising of this fluid through the system contribute to formation of pods, zones, and other rock units of unusual composition and texture. 3. Crystallization in the absence of silicate melt, yielding a wide variety of late-stage products. These include so-called "pocket minerals" and numerous mineral aggregates formed through exchanges of material among aqueous fluid and earlier-formed crystal-line phases. Development of pegmatite bodies can begin with either Step 1 or Step 2, but it is suggested that the processes involved in Step 2 are essential to the formation of all true pegmatites of igneous origin. The appearance of a second fluid phase, in general a supercritical aqueous fluid derived from the crystallizing melt, is regarded as the decisive event; it is promptly followed by fundamental changes in distribution and texture of the solid phases being formed. The processes can operate effectively in a fully closed system, and they also can modify the surrounding rocks if the system is open at any stage. Step 1 can include reactions between magma and earlier-formed crystals, but far more rapid and extensive exchanges of materials are subsequently effected by processes included in Steps 2 and 3; indeed, such exchanges also can account satisfactorily for pegmatites of metamorphic origin. Crystallization of most granitic magmas in the absence of a separate aqueous phase probably would begin within the temperature range 1,300°-650 ° C, the specific liquidus temperature depending mainly upon the amounts of volatile constituents held in solution at the time. This compositional factor also would be important in controlling the stage of crystallization-late, intermediate, or early-at which a separate aqueous fluid would make its appearance. Depending upon confining pressure as dictated by geologic conditions for a given system, the stage in crystallization represented by the presence of both silicate melt and aqueous fluid could begin within about the same temperature range of 1.300°-650 ° C. Exhaustion of the melt could occur within range extending downward to temperatures of 600C or even somewhat lower. Textural and structural features appear to be the most reliable indicators of the stages and fundamental processes involved in crystallization of both natural and synthetic pegmatites. The contrasting processes of crystallization from one fluid and from more than one fluid can operate over such broad P-T-X ranges that simple genetic pegmatite classifications based largely upon "key minerals," presumed temperature or pressure intervals, or the presence or absence of supercritical conditions appear to be somewhat unrealistic. © 1969 Society of Economic Geologists, Inc.