We review important advances in our understanding of the global carbon cycle since the publication of the IPCC AR4. We conclude that: the anthropogenic emissions of CO2 due to fossil fuel burning have increased up through 2008 at a rate near to the high end of the IPCC emission scenarios; there are contradictory analyses whether an increase in atmospheric fraction, that might indicate a declining sink strength of ocean and/or land, exists; methane emissions are increasing, possibly through enhanced natural emission from northern wetland, methane emissions from dry plants are negligible; old-growth forest take up more carbon than expected from ecological equilibrium reasoning; tropical forest also take up more carbon than previously thought, however, for the global budget to balance, this would imply a smaller uptake in the northern forest; the exchange fluxes between the atmosphere and ocean are increasingly better understood and bottom up and observation-based top down estimates are getting closer to each other; the North Atlantic and Southern ocean take up less CO2, but it is unclear whether this is part of the ‘natural’ decadal scale variability; large-scale fires and droughts, for instance in Amazonia, but also at Northern latitudes, have lead to significant decreases in carbon uptake on annual timescales; the extra uptake of CO2 stimulated by increased N-deposition is, from a greenhouse gas forcing perspective, counterbalanced by the related additional N2O emissions; the amount of carbon stored in permafrost areas appears much (two times) larger than previously thought; preservation of existing marine ecosystems could require a CO2 stabilization as low as 450 ppm; Dynamic Vegetation Models show a wide divergence for future carbon trajectories, uncertainty in the process description, lack of understanding of the CO2 fertilization effect and nitrogen–carbon interaction are major uncertainties.