In eastern Papua New Guinea, the seismicity, geomorphology, and results of seismic reflection profiling, drilling, and structural geological analysis reveal a zone of active low-angle normal faulting: the only clear known active example of this process worldwide. This Woodlark extensional province is ~250 km long in the east-west direction, located within the continental crust to the south of the Solomon Sea ocean basin. It adjoins the young oceanic spreading center farther east in the Woodlark Basin and since ~4 Ma has accommodated northward extension at rates of in excess of 10 mm yr-1 on low-angle normal faults that are rooted to the north. This geometry and the region's earlier history are assessed in the light of modern ideas for the mechanics of low-angle normal faulting to derive a working hypothesis for the existence of this process in this region. The extension in the Woodlark extensional province and oceanic spreading in the Woodlark Basin are thus interpreted as consequences of the slab pull force due to northward subduction of the Solomon Sea floor, which has caused the small Solomon Sea plate to separate from the Australian plate. It is suggested that this plate geometry developed by ~6--5 Ma following the ending of southward subduction of the Solomon Sea floor beneath what is now the Woodlark extensional province. The double thickness of lithosphere that formerly existed locally cooled the continental crust, making it brittle to the Moho. Following the end of this subduction, the base of the continental lithosphere became progressively exposed to the asthenosphere, causing dramatic heating of the overlying continental crust. This resulted in transient thermal disequilibrium, inducing transient horizontal lower crustal flow. The shear tractions imparted on the base of the brittle layer by such flow (combined with the stress components that develop after some of the lower crust has flowed from beneath a locality, requiring it to be supported laterally rather than from below) will rotate the principal axes of the stress tensor sufficiently to enable low-angle normal faults to form, and it is estimated that this sense of deformation was initiated around 4 Ma. The geometry of this study region means that this process will create north dipping low-angle normal faults, as are observed. This scheme can also account for a great deal of other observational evidence, including the available thermochronologic evidence and the observed vertical crustal motions. To explain this instance of active low-angle normal faulting thus requires knowledge of its relationship to earlier subduction and of effects of temperature-dependent changes to continental crustal rheology in crust that was initially cold.