A global 3-dimensional chemistry/transport model able to describe O₃, NO_x, Volatile Organic Compounds (VOC), sulphur and NH₃ chemistry has been extended to simulate the temporal and spatial distribution of primary and secondary carbonaceous aerosols in the troposphere focusing on Secondary Organic Aerosol (SOA) formation. A number of global simulations have been performed to determine a possible range of annual global SOA production and investigate uncertainties associated with the model results. The studied uncertainties in the SOA budget have been evaluated to be in decreasing importance: the potentially irreversible sticking of the semi-volatile compounds on aerosols, the enthalpy of vaporization of these compounds, the partitioning of SOA on non-carbonaceous aerosols, the conversion of aerosols from hydrophobic to hydrophilic, the emissions of primary carbonaceous aerosols, the chemical fate of the first generation products and finally the activity coefficient of the condensable species. The large uncertainties associated with the emissions of VOC and the adopted simplification of chemistry have not been investigated in this study. Although not all sources of uncertainties have been investigated, according to our calculations, the above factors within the experimental range of variations could result to an overall uncertainty of about a factor of 20 in the global SOA budget. The global annual SOA production from biogenic VOC might range from 2.5 to 44.5 Tg of organic matter per year, whereas that from anthropogenic VOC ranges from 0.05 to 2.62 Tg of organic matter per year. These estimates can be considered as a lower limit, since partitioning on coarse particles like nitrate, dust or sea-salt, together with the partitioning and the dissociation of the semi-volatile products in aerosol water has been neglected. Comparison of model results to observations, where available, shows a better agreement for the upper budget estimates than for the lower ones.