For over 6 million years in the Early Miocene, the Northland Peninsula of New Zealand's North Island was the site of a volcanic are, comprising north-westerly trending twin belts of marine and subaerial andesitic volcanoes. Approximately 50 volcanic edifices, from major massifs to small cones, have been newly identified on marine seismic profiles of the western belt, which is largely buried. The volcanoes were important point sources of sediment, and the volcaniclastic deposits of each of the major and intermediate sized volcanoes form distinct seismic sequences. The three main elements of the volcanic sequence are (i) a main body, (ii) a sloping apron and (iii) a wide, almost flat ring plain. The downlap of the oldest discernible internal clinoform reflector, relative to a known regional marker reflector underlying it, indicates the approximate time of onset of activity, whereas downlap of the top bounding reflector relative to those of other volcanic sequences provides a relative order of extinction. Sequence stratigraphic principles and superposition determine relative stratigraphic hierarchy. The volcanoes studied here were active only in the Early Miocene. With few exceptions, most were active by 22 Ma and extinct by 16 Ma, with many extinct by 19 Ma. There was no pattern in their order of inception or extinction, although a few became dominant, burying or surrounding their neighbours. At the south-eastern end of the chain there is a rapid change over a short distance in both age and orientation, from the northwest trending Early Miocene volcanism in the Northland Basin to north-east trending Middle and Late Miocene volcanism in the contiguous Taranaki Basin. The Northland volcanoes are interpreted to have been an are facing the south-westwards subducting South Fiji Basin, whereas the Taranaki volcanoes are interpreted to have been part of a separate, Pacific-facing volcanic are - the Lau-Colville Ridge.