The end of the Carboniferous, and the end of its second sub-period, the Pennsylvanian, is placed at about 299 million years ago. The supercontinent of Pangaea was mostly but not completely formed, and the glacial period that typified the following Permian Period was underway. The global climate was cooler and drier than it had been during most of the Pennsylvanian.
There is no global mass extinction to mark the end of the Carboniferous. The rainforest collapse was about 305 million years ago, six million years or so before the end of the Period.
So how is the end of the Pennsylvanian, the end of the Carboniferous, defined? This is another boundary like the end of the Silurian that’s in the category of “not with a bang.” Officially, the base of the Permian, and therefore the top of the Carboniferous, is placed below the first appearance of a particular conodont species. The internationally accepted location that defines it is in the Ural Mountains of Russia, where Permian rocks are well exposed.
All this is not exactly satisfying – there’s a tendency to want the major periods of earth history to be separated by major events. It might be a little more fun if the boundary was at the end of the rainforest collapse episode. But it’s not.
Despite the international agreement on the timing of the Pennsylvanian-Permian boundary, you will definitely see a diversity in references to the age of that boundary, sometimes indicating it’s at 280 million years ago and other times as well. Remember that geologic events occur at different times at different places – so even with international agreement, it’s reasonable to assign things like these boundaries to somewhat different times depending on where you are in the world. Saying the Pennsylvanian “ended” at 298.9 million years ago is just a human construction anyway.
—Richard I. Gibson