What kind of world dawned after the K/P boundary? We know from studies across localities in the USA that there is evidence of frequent wildfires continuing into the earliest Paleogene. But what happened to the atmospheric oxygen level after recovery from the K/P mass extinction—did it remain above modern levels? Were we still in a high-fire world? If there were fires, what is the evidence in the charcoal record, and do we know anything about the vegetation that was burning? When the charcoal in the coal database was originally compiled, one of the important issues was how we recorded and represented our data. Early to mid-Paleocene Epoch coals (from around 65 to 55 million years ago) are often recorded as ‘earliest Tertiary’ in coal literature. (The Tertiary was the name we used to use for what we now call the Paleogene and Neogene Periods, stretching from around 65 to 1 million years ago.) However, coals that are nearer to the start of the Eocene Epoch, just older than 55 million years ago, are notoriously difficult to date. This is a problem we have with many coal sequences, as they are deposited on land, and most of the fossils used to give ages are found in marine waters. Many coals of this age are often simply recorded as coming from the late Paleocene or early Eocene. Where we have good dating information, Paleocene coals all tend to have high inertinite (charcoal) contents, well above 19 per cent. By the mid to late Eocene (50–40 million years ago), however, worldwide the charcoal contents are low, around 5 per cent or even less. There must, therefore, have been a fundamental change in the Earth system at this time. Another problem is the way in which we chose to represent our data and show the calculated oxygen curve. In order to get sufficient data to plot the curves we decided to use 10-millionyear bins. This was not a problem for the Paleozoic–Mesozoic transition, covering the great Permian mass extinction, which took place 250 million years ago.