Evaluation of the extent of high-level contamination in granites requires a determination of the amount of restite or resistate (resister) present. High-level S-type granites of the Lachlan Fold Belt (LFB) of southeastern Australia are the most consistently interpreted in terms of the "restite unmixing hypothesis." Therefore, they should provide the clearest evidence for restite in high-level plutons; however, microstructural and isotopic evidence either is against restite or is ambiguous. Metasedimentary enclaves appear to be accidental xenoliths collected from metamorphic rocks above unexposed granulite-facies source-rocks; microgranitoid (orthopyroxene microtonalite) enclaves have igneous microstructures and show structural and isotopic evidence of mixing of S-type felsic magma with more mafic magma before becoming mingled to form enclaves in the host pluton. The origin of the single grains and small aggregates that have been considered to be the main evidence for restite in the S-type granites of southeastern Australia is difficult to determine. Unobserved deep, low-pressure and high-temperature granulite-facies sources (possibly varieties of "MASH" zones), rather than typical migmatite complexes, may be applicable for large, high-level S-type granites. A possible example of the upper part of such a melting zone is the Hidaka Metamorphic Belt, Hokkaido, Japan.