This chapter summarizes the factors that determine future CO2 concentration changes, quantifies uncertainties in concentration projections, and assesses the importance of these uncertainties in predicting future climate change. The 1992 IPCC emissions scenarios are used to determine future changes in CO2 concentration, radiative forcing and global-mean temperature, and uncertainties accruing at each calculation stage are estimated. The main factor determining temperature change uncertainty is the climate sensitivity (∆T2x); uncertainties due to this factor are some 3-5 times larger than uncertainties arising from carbon cycle modelling uncertainties. Uncertainties arising from ∆T2x are always larger than those due to uncertainties in emissions. This is particularly so over the next 50 years, during which time emissions-related uncertainties in global-mean temperature change are small. By 2100, ∆T2xand emissions-related uncertainties become comparable. The relative importance of carbon cycle and emissions-related uncertainties also depends on time, with the former being dominant to about 2040. Carbon cycle uncertainties are largely controlled by uncertainties in the magnitude of the so-called “missing sink”. In terms of research priorities, it is not the associated uncertainties that should be the determining factor, but the potential for (and cost-effectiveness of) reducing these uncertainties. The greatest scope for uncertainty reduction, at least in the short term, appears to lie in the area of carbon cycle modelling.