Chemical analyses of 167 typical specimens indicate that about 95 percent of the intrusive rocks of the central Sierra Nevada contain more than 79 percent normative Ab + An + Or + Qz. If the composition of the lower continental crust is similar to or slightly more felsic than andesite, as seems likely, the system NaAlSi3O8-CaAl2Si2O8-KAlSi3O8-SiO2-H2O provides an excellent chemical model for testing various schemes of fusion of the lower crust and crystallization of the resulting magmas. From consideration of this system in conjunction with field and petrographic data, we conclude that the intrusive rocks are best explained by repeated episodes of equilibrium fusion corresponding to magmatic sequences defined by field, petrologic, chemical, and geochronologic data. Fractional crystallization of the crystal-liquid mush generated by equilibrium fusion, coupled with periodic upward or lateral movement of the less crystallized central part of the magma, would produce the characteristic mafic to felsic sequence of intrusion; each mafic to felsic sequence corresponds to a separate equilibrium fusion event. In contrast, a close approach to fractional fusion of the lower crust is inadequate for obtaining most of the plutonic rocks, because rock compositions capable of being produced by this process do not match those observed. Normal amounts of conductive heat from the mantle and from radioactive decay in the crust may have been capable of causing fusion in the deepest parts of a thickened crust under the central part of the Sierra Nevada without the aid of a transient heat source from the mantle, but would have been inadequate where the crust was thin in the western Sierra Nevada. However, upward transport of andesitic and basaltic magmas generated along a Mesozoic subduction zone dipping beneath the Sierra Nevada would have provided sufficient additional heat to make fusion of the lower crust unavoidable. This implies that a major portion of the present batholith must have been derived from the lower crust.