In this paper, we carry out a meta-analysis of recent studies into the costs of greenhouse gas mitigation policies that aim at the long-term stabilisation of these gases in the atmosphere. We find the cost estimates of the studies to be sensitive to the stringency of the stabilisation target, the assumed emissions baseline, the way in which the time profile of emissions is determined in the model, the choice of control variable (CO2 only versus multigas), the number of regions and energy sources in the model and, to a lesser degree, the scientific "forum" in which the study was developed. We find that marginal abatement costs of the stringent long-term targets that are currently considered by the European Commission are still very uncertain but might exceed the costs that have been suggested by recent policy assessments. (C) 2008 Elsevier Ltd. All rights reserved.