Widespread cropland abandonment occurred after the collapse of socialism across the former Soviet Union, but the rates and spatial patterns of abandoned lands are not well known. As a result, the potential of this region to contribute to global food production and estimates of the carbon sink developing on currently idle lands are highly uncertain. We developed a spatial allocation model that distributes yearly and sub-national sown area statistics to the most agriculturally suitable plots. This approach resulted in new, high-resolution (1 km2) annual time series of cropland and abandoned lands in European Russia, Ukraine and Belarus from 1990 to 2009. A quantitative validation of the cropland map confirms the reliability of this dataset, especially for the most important agricultural areas of the study region. Overall, we found a total of 87 Mha of cropland and 31 Mha of abandoned cropland in European Russia, Ukraine and Belarus combined, suggesting that abandonment has been severely underestimated in the past. The abandonment rates were highest in European Russia. Feeding our new map dataset into the dynamic vegetation model LPJmL revealed that cropland abandonment resulted in a net carbon sink of 470 TgC for 1990 to 2009. Carbon sequestration was generally slow in the early years after abandonment, but carbon uptake increased significantly after approximately 10 years. Recultivation of older abandoned lands would be associated with high carbon emissions and lead to substantial amounts of carbon not being sequestered in vegetation formations currently developing on idle croplands. Our spatially and temporally explicit cropland abandonment data improve the estimation of trade-offs involved in reclaiming abandoned croplands and thus in increasing agricultural production in this globally important agricultural region.