The multiply deformed (D1–D3) Palaeoproterozoic Willyama Supergroup in south–central Australia incorporates upper and lower c. 1700 Ma metasedimentary sequences with contrasting early tectonothermal histories that invite comparisons with the metamorphic core complexes and younger extensional orogens of western North America and Europe. A detachment surface of D1 age separating these two sequences has the deduced geometry of an extensional shear zone, juxtaposing rocks subjected to bimodal magmatism, sillimanite- to granulite-grade migmatization, and Na–Fe metasomatism against a less intensely metamorphosed upper plate lacking both migmatites and bimodal magmatism. Synextensional metamorphism took place under low-pressure–high-temperature conditions, producing regionally extensive andalusite- and sillimanite-bearing mineral assemblages before further high-grade metamorphism accompanying D2 recumbent folding and crustal thickening. D2 folding locally inverted the original D1 thermal structure so that sillimanite-grade lower plate rocks now lie structurally above andalusite-grade rocks of the upper plate, rendering recognition of the original detachment surface and associated thermal structure difficult. U–Pb dating of synextensional metabasites intruded into lower plate rocks just below the detachment surface indicate that extension and related bimodal magmatism peaked around 1690–1670 Ma, some 70–90 Ma earlier than the 1600–1590 Ma age previously inferred for the onset of regional deformation and low-P–high-T metamorphism. A regionally extensive redox boundary associated with the detachment surface served as the locus for several important mineral deposits.