The Okataina Volcanic Centre (OVC) is a highly active modern silicic magmatic system at the northeastern end of the central Taupo Volcanic Zone, North Island, New Zealand. This study presents a set of petrographic, mineralogical, geochemical and isotopic data that focuses on the rhyolite lavas erupted from the OVC. These data have been combined with existing geochronological data and knowledge of vent positions to identify magma types, and hence provide a model for the spatial and temporal evolution of the silicic magmatic system. Representative samples of OVC rhyolite lavas were analysed for whole rock major and trace elements by XRF (90 samples), trace and rare earth elements by LA-ICPMS (32), and Sr (25), Nd (8) and Pb (4) isotopes by TIMS. Major element compositions of phenocrysts in 24 representative lavas were determined by EPMA. Rhyolite lavas of the OVC contain phenocrysts (5 - 35 %) of plagioclase, quartz, Fe-Ti oxides ± orthopyroxene ± calcic amphibole ± cummingtonite ± biotite. Apatite and zircon are accessory. Temperatures determined from Fe-Ti oxide geothermometry range from ∼ 650 - 830°C. The lavas are dominantly peraluminous, medium-K, calc-alkaline rhyolites, with SiO₂ ≈ 72.5 - 77.5 wt. %, ⁸⁷Sr/⁸⁶Sr ≈ 0.70516 - 0.70583, εNd ≈ 0.02 - 1.70, ²⁰⁶Pb/²⁰⁴Pb ≈ 18.75 - 18.84, ²⁰⁷Pb/²⁰⁴Pb ≈ 15.55 - 15.64 and ²⁰⁸Pb/²⁰⁴Pb ≈ 38.42 - 38.74. The volcanic history of the OVC can be subdivided into three periods, (i) > 220 ± 10 ka, (ii) 220 ± 10 - c. 65 ka, and (iii) < c. 65 ka. The two earlier periods involved eruption of caldera-forming ignimbrites and rhyolite lavas. The most recent period of activity involved eruption of the caldera-modifying Mangaone Pyroclastics Subgroup (c. 43 000 - 31 400 years B.P.), and the caldera-infilling rhyolite lavas and pyroclastics comprising the Haroharo, Okareka and Tarawera volcanic complexes (c. 25 000 years B.P. - present). In each period, a number of rhyolitic magma types or batches have been identified and distinguished based on their mineralogical, geochemical and isotopic characteristics. At least nine magma batches may have been erupted in the last 25 000 years, with estimated maximum generation times from c. 3 800 - 15 500 years, minimum generation rates from 3.1 x 10⁻⁴ - 3.2 x 10⁻³ km³/year, residence times from c. 4 000 - 31 000 years and erupted volumes generally < 15 km³. The Haroharo Volcanic Complex was built up in four rhyolitic eruptive episodes beginning c. 25 000 years B.P. The lavas and pyroclastics erupted in the Te Rere episode were derived from a single, genetically discrete, relatively volumetrically small magma batch. The lavas and pyroclastics erupted in the Rotoma, Mamaku and Whakatane episodes record the evolution of a relatively larger single magma batch over c. 4 000 years by closed-system fractional crystallisation processes. The Okareka Volcanic Complex was built up in two rhyolitic eruptive episodes at c. 25 000 and 15 800 years B.P., and a genetic relationship is suggested between the magma erupted in these episodes. The Tarawera Volcanic Complex was built up in four rhyolitic eruptive episodes beginning c. 22 500 years B.P. Multiple magma batches were erupted during the Okareka and Rerewhakaaitu episodes. The lavas and pyroclastics erupted in each of the two youngest episodes (Waiohau and Kaharoa) were derived from single, genetically discrete magma batches. The eruption of multiple, relatively small volume, genetically discrete magmas in the last c. 25 000 years precludes the coalescence of a large magma body beneath the OVC, and has implications for volcanic hazard assessment and eruption prediction. Future eruptions may involve either residual silicic magmas revived by basaltic intrusion, or basaltic magmas that have reached the surface without interacting with residual silicic magmas because they have crystallised, or new genetically discrete silicic magma. Mineralogical, geochemical and isotopic data obtained for rhyolite lavas erupted from the adjacent Rotorua and Kapenga volcanic centres reveal that their evolution has also involved eruption of multiple magma batches, but lack of geochemical, isotopic and age data preclude the development of any spatial and temporal models for these centres.