Isotopic mapping reveals the location of crustal fragments along a long-lived convergent plate boundary

New Guinea has acted as the boundary between the Australian and Pacific plates for hundreds of millions of years. Strike-slip movement and arc–continent collisions along this boundary during the Cenozoic have shuffled rocks of different age and composition in a series of terranes along the plate boundary making mapping them a considerable challenge. Here we report results of SrNd isotopic data obtained from rock samples from western New Guinea that are representative of the different terranes. These isotopic data reveal the crustal affinity of the terranes and we have used these data to map their spatial distribution. The isotopic data show three distinct crustal domains underlying western New Guinea; Palaeozic–Mesozoic Australian continental crust (87Sr/86Sr = 0.719594 to 0.710921; εNd = −13.85 to 1.373); thinned transitional crust intruded by Miocene–Pleistocene magmatic rocks (87Sr/86Sr = 0.706524 to 0.704019; εNd = 6.67 to 2.13); and accreted island arc crust (87Sr/86Sr = 0.704053 to 0.703759; εNd = 6.63 to 4.97). These data, together with crustal contamination models, indicate that the northern-most extent of Australian continental crust exists beneath the northern-most section of western New Guinea. We also combined our isotopic data with existing data across New Guinea and used these to develop an isotopic map that shows the position of the ancient Australian–Pacific Plate boundary, producing results that are also consistent with broad-scale seismic tomography imagery. Our findings provide a framework for mapping other plate boundaries, particularly ancient systems where only fragmentary data exist.