We present an integrated record of environmental change from Lake Omapere, Northland, New Zealand, based on palaeolimnological analysis of a 7-m-long core and eight adjunct cores spanning part of the last 80 calendar (cal.) ka. The chronology was developed using tephrochronology, palaeomagnetism and radiocarbon dating together with climato- and palyno-stratigraphy. Two of 14 tephra layers in the cores provide markers for correlating the record with other New Zealand climato-stratigraphic sequences, deep-sea cores and the global marine isotope record. Pollen, diatom, palaeomagnetic, sedimentologic and pigment analyses show that Lake Omapere, currently 2 m deep, has had a discontinuous history. Occupying a shallow basin perched partly within old basalt lavas, the initial (alkaline) lake formed towards the end of MIS 5a at ca. 80 cal. ka, presumably because of blockage of local drainage of a swampy alluvial floodplain, inundating peat deposits and forest trees, including kauri (Agathis australis). The lake filled rapidly to a level 1–2 m above that at present. Such filling was probably unrelated to climate but a subsequent phase of variable but generally falling lake levels and increasing dystrophy may have been climatically controlled, commencing during MIS 4 and culminating in periods early in MIS 3 when the lake became swampy or dry. Conditions of non-deposition (or non-preservation) obtained for most of the period after ca. 55 cal. ka (Rotoehu Ash) until formation of the modern lake approximately 600–700 cal. years ago or soon after, as indicated by the presence of Tephra-1, identified in part as Kaharoa Tephra (AD 1314), near the top of the core. The modern lake originated possibly through damming of the western outlet as a consequence of accelerated erosion accompanying earliest Polynesian deforestation, an interpretation supported by Maori oral tradition. The pollen record indicates that beech Nothofagus (presumably N. truncata) was much more common in Northland during the Last Glacial, and that for this region the relative abundance of Nothofagus vs. Agathis pollen serves as a better indicator of cooler versus warmer intervals during the Quaternary than the ratio of tree to non-tree pollen. However, it seems likely that moisture balance was a more critical factor than temperature in vegetation composition and distribution, particularly during the LGM, and the long periods of hiatus may also be linked to a drier climate than present. Correlation coefficients between the pollen curves confirm that several tree species (Halocarpus bidwillii, H. biformis and Phyllocladus alpinus), previously palynologically concealed within generic taxonomic groups, occurred about two degrees latitude further north and at a much lower altitude than their current limits during cooler or drier phases of the Last Glacial. A temperature depression in Northland of 4 °C at various times during the Last Glacial is inferred from these range expansions. Nevertheless, the persistence of widespread forest cover suggests that the late Pleistocene climate of Northland was less severe than for most of the rest of New Zealand and strengthens the argument for a heightened temperature gradient across northern New Zealand during the Last Glacial.