With the widespread usage of multi-region multi-sector CGE models in various areas of policy analysis, there is an increasing need to provide a validation of the commonly employed baseline scenario. So far, there have only been a few attempts to validate CGE model outcomes in a systematic way. This study is a first attempt to validate a multi-country, multi-sector CGE model. The paper broadly follows the ‘historical’ approach to validate the MAGNET – the GTAP based model of LEI-Wageningen University. The GTAP 6 database with base year 2001 is used to construct projections for the period 2001-2007, which are subsequently compared with historical time-series for output, value added and labour for a large number of countries and detailed sectors. Finally, regression analysis is applied to investigate if forecast performance structurally differs across sectors and countries. This information can be used as a guidance to improve modelling.