It is generally accepted that oxygen in the atmosphere rose in two major steps at s.round 2.4-2.2 and 0.7-0.5 billion years ago. The variation in atmosphere oxygen over the last 500 million years, is considered to have been relatively minor by comparison. Sedimentary pyrite from marine shales efficiently captures many trace elements from the oceans, providing a novel proxy for seawater chemistry. Here we use temporal changes in the selenium and cobalt content of Phanerozoic marine pyrite, coupled with the 87Sr/86Sr ratio in marine carbonate, to argue for five dramatic p02 cycles, each starting with a period of oxygenation, followed by a period of de-oxygenation. The selenium proxy is based on the premise that increased erosion of continental rocks leads to the release of selenium as both the selenate and selenite species. Under neutral to alkaline, oxygenated conditions the selenate species remains highly soluble, where it can be readily transported via river systems to the ocean. Cobalt on the other hand becomes less soluble under increasing p02 as the oxidized species Co2 and CoO are immobilised by Fe and Mn oxyhydroxides, that form during weathering. Thus variations in the Se and Co composition of marine pyrite enable us to propose a new oxygenation proxy; the ratio Se/Co, which increases in marine pyrite during periods of increasing p02 (oxygenation) and decreases during periods of decreasing p02 (deoxygenation). The first half of each of the five Phanerozoic p02 cycles involves an increase in atmosphere/ocean oxygenation driven initially by supercontinentdispersal, increased continental erosion and nutrient trace element flux to the oceans. Increased marine productivity leads to carbon and sulphur sequestration, producing metalliferous black shales, and further drives oxygenation to the peak of the cycle. The cycle downside suggests decreasing oxidative erosion and nutrient delivery, resulting in a drop in productivity. Continued drawdown of ocean trace elements leads to a productivity crisis and widespread depletion of trace elements in the oceans, ultimately contributing to three of the five major Phanerozoic marine mass extinction events. Continental erosion rates and trace element nutrient flux are suggested as the prime drivers of the cycles.