For as long as agriculture has existed, agricultural land abandonment (ALA) has been a globally relevant land use change. Depending on the timescale considered and the definitions and methods used, spatial estimates of historical ALA range in the several hundreds of millions of hectares. ALA implies the spontaneous recovery of ecosystem properties towards pre-disturbance states. Because agricultural lands are often degraded and carbon depleted, the natural ability of abandoned agricultural lands to act as carbon sinks has been, and will continue to be, a significant component of the terrestrial carbon cycle. Here, we provide a brief snapshot of the history of ALA, its drivers, and its known ecosystem carbon impacts from ancient times to the present, especially since the mid-20th century. We then explore the current and future implications of ALA-derived carbon sequestration in Europe, focussing on soil organic carbon based on synthesized published data (chronosequences and paired plots) and land surface model estimates. The majority of abandoned agricultural lands serve as carbon sinks, but there are clear scenarios where carbon may be lost or unchanged even after several years post-agriculture. Our results show that management of abandoned agricultural lands must consider multiple factors such as past land use practices (e.g., croplands vs pastures, past crop types, etc.), future land use management practices (e.g., natural or assisted restoration), local climate variables, and the present soil quality and carbon stock to ensure steady carbon sequestration following agricultural cessation. To avoid lost opportunities for climate change mitigation, ALA requires dedicated research and policy attention because: 1) it is a widespread, ongoing global land use change; 2) it does not always result in carbon sequestration; 3) its carbon gains are often lost in the first few decades when agriculture is re-established; 4) and it can facilitate wildfires which can also reverse carbon gains.