In bone tissue engineering, both geometrical and mechanical properties of a scaffold play a major part in the success of the treatment. The mechanical stresses and strains that act on cells on a scaffold in a physiological environment are a determining factor on the subsequent tissue formation. Computational models are often used to simulate the effect of changes of internal architectures and external loads applied to the scaffold in order to optimise the scaffold geometry for the prospective implantation site. Finite element analysis (FEA) based on computer models of the scaffold is a common technique, but would not take into account actual inaccuracies due to the manufacturing process. Image based FEA using CT scans of fabricated scaffolds can provide a more accurate analysis of the scaffold, and was used in this work in order to accurately simulate and predict the mechanical performance of bone tissue engineering scaffolds, fabricated using selective laser sintering (SLS), with a view to generating a methodology that could be used to optimise scaffold design. The present work revealed that an approach that assumes isotropic properties of SLS fabricated scaffolds will lead to inaccurate predictions of the FE model. However, a dependency of the grey value of the CT scans and the mechanical properties was discovered, which may ultimately lead to accurate FE models without the need of experimental validation.