Computable general equilibrium (CGE) modelling represents a powerful tool for hypothesising possible sustainability outcomes that might be triggered with the implementation of policy proposals. Nevertheless, CGE modelling is based on several tight general equilibrium and neoclassical micro-economic theoretical assumptions that make their application to the assessment of all three pillars of sustainability questionable. Although some of these assumptions have been relaxed in recent and more advanced CGE models, further research needs to be undertaken in order to bring model specifications closer to realistic behavioural relationships. CGE models also tend to focus on alternative equilibrium outcomes and rarely deal with the adjustment process or regulation measures needed to realistically bring the economy into the desired new equilibrium stance. Moreover, CGE models inherently face severe rigidities when attempting to deal with environmental and social effects. However, some authors have argued that CGE modelling may provide a suitable backbone for all three dimensions of Sustainability Impact Assessment (SIA). The paper takes a critical stand and supports the view that though CGE models may provide some useful information on individual, particularly economic, impact aspects of policy reforms, it may be inappropriate and even misleading to rely extensively on their use in SIAs.