Analysis of the structure and sedimentary geology of western New Zealand has identified a middle Eocene to early Miocene continental rift system, 1,200 km long and 100–200 km wide, named here the “Challenger Rift System.” Four phases of rift development occurred: (1) infra-rift subsidence, (2) active axial trough subsidence, (3) expanded rift subsidence involving collapse of the rift shoulders, and (4) incipient sea-floor spreading. The spatial and temporal distribution of these phases identifies a North Island and a South Island rift segment and shows that rifting propagated toward the center of the rift from both ends. The northern segment shows a simple pattern of rifting that is comparable with Vink's model of rift propagation; the southern segment, with locked zones and rift nucleation segments, is comparable to Courtillot's model of rift propagation. The sea-floor–spreading history of the southwest Pacific shows that the northern rift segment probably linked with a sea-floor–spreading center in the Norfolk Basin, and the southern segment linked with the Southeast Indian Ridge. This is corroborated by the good correlation between the ages of sea-floor magnetic anomaly lineations that are aligned with the rift and the biostratigraphic ages of rifting.
The probable continuity of the rift system in its early development precludes pre-Miocene transcurrent displacement on the Alpine fault; an early Miocene (23 m.y. B.P.) age of Alpine fault inception is indicated by the age and pattern of rift disruption attributed to compression that originated at the Australia–Pacific plate boundary. The modern Australia and Pacific plates were not discrete entities, therefore, until the early Miocene.