When electromagnetic waves propagate through wood they will be affected by variations in the dielectric properties. These variations are related to different properties within the wood. Dielectric properties of wood depend on moisture content, density, temperature, grain angle and frequency of the electromagnetic wave. The aim with this study was to use threedimensional finite element modelling (FEM) in combination with multivariate statistics and projection to latent structures (PLS) to develop a prediction model for moisture content and dry basis density in wood. An X-ray computed tomography (CT) scanner was used to detect density and, with that, moisture content in wood. Dielectric properties were calculated within a wood volume and used in the simulation. Different frequencies and directions of the electric field were simulated. Previous studies have shown that this method gives a good description of the interactions between wood and microwaves in two dimensions. The results show that 3D finite element modelling can be used to generate a prediction model for moisture content from microwave scanning. Other properties of wood such as density and grain angle have less influence on the microwaves but might be predicted by further processing of the signal.