The potential for feedbacks between terrestrial vegetation, climate, and the atmospheric CO2 partial pressure have been addressed by modelling. Previous research has established that under global warming and CO2 enrichment, the stomatal conductance of vegetation tends to decrease, causing a warming effect on top of the driving change in greenhouse warming. At the global scale, this positive feedback is ultimately changed to a negative feedback through changes in vegetation structure. In spatial terms this structural feedback has a variable geographical pattern in terms of magnitude and sign. At high latitudes, increases in vegetation leaf area index (LAI) and vegetation height cause a positive feedback, and warming through reductions in the winter snow-cover albedo. At lower latitudes when vegetation becomes more sparse with warming, the higher albedo of the underlying soil leads to cooling. However, the largest area effects are of negative feedbacks caused by increased evaporative cooling with increasing LAI. These effects do not include feedbacks on the atmospheric CO2 concentration, through changes in the carbon cycle of the vegetation. Modelling experiments, with biogeochemical, physiological and structural feedbacks on atmospheric CO2, but with no changes in precipitation, ocean activity or sea ice formation, have shown that a consequence of the CO2 fertilization effect on vegetation will be a reduction of atmospheric CO2 concentration, in the order of 12% by the year 2100 and a reduced global warming by 0.7°C, in a total greenhouse warming of 3.9°C.