The Hadean (before 4.0 Ga) crust has long been considered to comprise mainly primitive mafic and ultramafic rocks. However, mineral inclusions in detrital zircons as old as 4.4 Ga from Jack Hills, Australia, have been interpreted to be magmatic and to provide evidence for extensive granitic crust. In situ U-Pb dating of monazite and xenotime inclusions in 4.25–3.35 Ga detrital zircons from Jack Hills shows that these inclusions are not magmatic, but formed during metamorphism at either 2.68 Ga or 0.8 Ga. Monazite-xenotime thermometry of intergrowths in the inclusions and the quartz-muscovite rock matrix constrain temperatures to between 420–475 °C, corresponding with conditions during peak regional metamorphism. Petrography and U-Pb geochronology of zircon inclusions from other localities show that the replacement of primary inclusions may commence in the igneous host rock and continue through weathering, sedimentation, and diagenesis. With increasing metamorphic grade, the inclusion assemblage increasingly reflects the composition of the rock matrix. In Jack Hills, most of the inclusions have the same composition and abundances as the metamorphic matrix, consistent with their formation during metamorphism. The titanium content of quartz inclusions indicates formation temperatures of 350–490 °C, supporting a metamorphic origin. Several lines of evidence indicate that at least some of the muscovite inclusions are also secondary. The lack of apatite inclusions in zircons from Jack Hills, relative to zircon in common granitic rocks, suggests that secondary minerals may have replaced primary apatite. Thus, detrital zircon may not be impermeable to post-depositional fluids, raising doubts about the use of the mineral inclusions they contain to infer initial magma chemistry. These results call for a reassessment of the source melts of the Hadean zircons and the composition of the earliest crust.