At Tarandore Point, 328 km south of Sydney, New South Wales, Australia, two tonalite phases are in contact with a gabbroic diorite, all belonging to the Moruya Batholith. Lobate to crenulate contacts, pillowing of the more mafic phase in the more felsic phase, net-veining of the more mafic phase by the more felsic, local chilling of the more mafic phase against the more felsic, and an abundance of rounded to elongate microgranitoid enclaves all suggest mingling of magmas at the site of emplacement. Therefore, all phases were emplaced before the others solidified, and an order of intrusion cannot be determined. The variety of microgranitoid enclaves suggests that repeated magma mingling has occurred. Linear geochemical trends indicate either magma mixing or restite unmixing as a prime cause of variation in the members of the Moruya suite of granitoids. Direct evidence of mixing is not present, but evidence that the magmas were substantially liquid at the time of emplacement sugests that mixing is possible. Restite unmixing is unlikely to be the prime process, because near-minimum melts, which would occur in all granitoid phases according to that model, should mix easily, rather than mingling. In fact, mingling has occurred between tonalite phases of very similar composition, which would not be expected if the liquid parts of enclaves were minimum melts. Available experimental and theoretical data on mixing in both magma chambers and conduits cannot be applied directly to the occurrence at Tarandore Point, owing to inadequately constrained parameters, but mixing generally should be favoured by higher temperatures, which is consistent with our inference that relatively high proportions of melt were present in the granitoid phases at the time of emplacement. Conduit mixing may be reflected in a local zone of strong to extreme elongation of enclaves, in which the distortion has been accomplished by magmatic flow without solid-state deformation, as evidenced by strong dimensional preferred orientations, coupled with a lack of microstructural evidence of intragranular plastic strain. This suggests that elongate enclaves should not be used as indicators of solid-state deformation or as tectonic strain markers without microstructural evidence that suitably strong intracrystalline plastic deformation has taken place.