Popular accounts often say that oxygen reached about 30 or 35 per cent of the atmosphere during the Carboniferous, compared with about 21 per cent today. The broad idea of a major late Palaeozoic oxygen rise is supported by geological evidence, but an exact percentage is not a direct measurement from a Carboniferous air sample. It comes from reconstructions that combine different kinds of evidence and assumptions.
The Carboniferous lasted from roughly 359 to 299 million years ago. Its rocks preserve extensive coal deposits, fossil charcoal, plants and animals from landscapes that included vast wetlands. These records help researchers investigate how the long-term carbon cycle changed the atmosphere. They do not show that every Carboniferous habitat was a swamp or that oxygen had one uniform value throughout the period.
How oxygen enters the geological record
Earth's atmosphere has no long-lived geological storage mechanism for free oxygen. Over long intervals, oxygen levels reflect a balance between sources and sinks. Photosynthesis releases oxygen, but most of that oxygen is consumed again as organic matter decays or reacts with minerals. When some organic carbon escapes oxidation and is buried in sediments, the oxygen produced during photosynthesis is not all used up again. Burial of pyrite also affects the balance.
Researchers therefore use models of the carbon and sulphur cycles to estimate ancient atmospheric oxygen. These models combine records such as carbon isotopes in sedimentary rocks with estimates of weathering, burial and recycling. Each output is a reconstruction based on a stated model, not a preserved sample of the atmosphere. Different choices about geological processes and proxy data can produce different curves.
Coal seams are evidence that plant material accumulated and was buried in particular environments. They are not themselves oxygen gauges. Their formation depended on vegetation, waterlogged landscapes, sediment supply, subsidence, climate and preservation. A coal deposit supports an account of carbon burial, but it cannot by itself say that the global atmosphere contained a specific percentage of oxygen.
What charcoal can and cannot establish
Fossil charcoal, also called inertinite in many sedimentary contexts, records ancient fires. Wildfire requires a minimum oxygen concentration, so the oldest securely identified charcoal provides a threshold: oxygen had reached a level at which combustion could occur. Abundant charcoal through a rock sequence can add information about fire regimes and fuel, but the signal also depends on vegetation, moisture, ignition and the chance that charcoal was preserved.
Charcoal is therefore a constraint on atmospheric oxygen, not a precision meter. A decline in charcoal may reflect a change in oxygen, fuel, climate or preservation. Researchers compare the charcoal record with isotope-based models and other geological evidence instead of reading a single curve as a direct observation.
Why estimates of a Carboniferous maximum differ
Published reconstructions agree more readily on broad change than on every peak and trough. One study may infer a high late Palaeozoic maximum from a particular isotope model; another may favour a lower peak or place a rise at a different time. Models can disagree because they handle sulphur isotope fractionation, organic-carbon burial, volcanic inputs and weathering differently. Their curves also depend on the quality and age control of the rock record.
A revised model of Paleozoic oxygenation compared its calculations with geochemical proxies and found important differences from an earlier widely used model, especially for early Paleozoic conditions. The authors supported a stepwise oxygenation history, while showing why the model itself matters. The lesson for the Carboniferous is not that every estimate is equally plausible: it is that numerical claims should name the evidence and model behind them.
The atmosphere and oceans must also be kept distinct. Atmospheric oxygen can rise while parts of the ocean remain oxygen-poor. Marine sediments record local and regional water conditions, which can vary with depth, circulation and biological productivity. A claim about oxygen in the air cannot automatically be applied to every ancient sea.
Did extra oxygen create giant insects?
The Carboniferous and Permian are famous for very large arthropods, including griffinfly relatives such as Meganeura and Meganeuropsis. High oxygen has often been proposed as one reason some insects could attain large sizes, because insects deliver oxygen through a branching tracheal system rather than a lung-and-blood system. The timing of large fossil insects and modelled oxygen highs makes the hypothesis worth testing, but coincidence alone does not establish cause.
A 2026 study tested a specific part of the explanation by measuring tracheoles in the flight muscles of 44 living insect species across ten orders, then extending the relationship to the giant Permian insect Meganeuropsis permiana. The tracheoles occupied about one per cent or less of the muscle volume in most species, and the researchers concluded that oxygen diffusion through this system does not set the maximum body size of living or giant prehistoric insects. This result challenges a simple oxygen-bottleneck explanation. It does not show that atmospheric oxygen never affected insect physiology, nor does it determine the oxygen content of Carboniferous air.
Insect size also varied across lineages and time. Flight mechanics, development, ecology, predators, available food and the fossil record all matter when researchers explain that pattern. The deep-time collection brings together further guides to the varied organisms and environments represented by these fossils.
A careful reading of the evidence
The geological record supports substantial changes in atmospheric oxygen through deep time and gives good reason to investigate a late Palaeozoic rise. Charcoal supplies a combustion threshold, coal-bearing strata document carbon-rich wetland deposits, and isotope-based models test how carbon and sulphur burial could alter atmospheric composition. Each line answers a different question and has different uncertainties.
For that reason, it is more accurate to describe a modelled Carboniferous oxygen high than to quote one exact percentage as a settled measurement. The Palaeozoic Era overview places the period in its wider geological context, while the geological time scale explains how period boundaries are defined. This collection on life in the distant past brings together other evidence-based guides to ancient environments.
Frequently asked questions
Was atmospheric oxygen definitely higher in the Carboniferous?
Geological models and proxy evidence support a major late Palaeozoic oxygen rise, but they do not produce one uncontested value for the entire Carboniferous. The exact timing and height of a maximum remain dependent on the reconstruction used.
How can fossil charcoal provide evidence about oxygen?
Combustion needs a minimum oxygen concentration, so charcoal shows that the atmosphere exceeded a threshold. Charcoal abundance also depends on fuel, moisture, ignition and preservation, which means it is not a precise oxygen gauge.
Did coal-forming forests directly raise atmospheric oxygen?
Photosynthesis releases oxygen, and long-term burial of some organic carbon prevents all of that oxygen from being consumed again. Coal deposits record particular burial settings, but a coal seam alone cannot quantify global oxygen.
Did higher oxygen cause Carboniferous insects to become giant?
It remains a proposed influence, not a complete explanation. A 2026 study challenged the idea that oxygen diffusion through flight-muscle tracheoles sets the maximum insect body size, and other ecological and biomechanical factors also need consideration.