In this paper we use the novel Disturbed Stress Field Model (DSFM) from Facconi et al. (2013) to model the lateral in-plane behaviour of different unreinforced masonry (URM) walls which we tested at the laboratory of the EPF Lausanne, Switzerland. The model uses the failure criterion developed by Ganz (1985) for URM and is implemented in the software VecTor 2. In the first part of this article, we show that the DSFM gives good estimates for the initial stiffness of the walls but underestimates the displacement and force capacity of our URM walls. This is due to the confinement of the mortar base joint which is confined by the wall foundation. However, we show that this phenomenon can be accounted for by replacing the compression strength of the masonry of the first brick layer by the strength of the brick itself. Using this improved model, we compare the numerical results to the experimental results with regard to displacements and strains. We show that the simplification of the masonry to a continuous material is correct when comparing the global engineering demand parameters (EDPs), e.g., force-displacement behaviour of the walls or crack pattern. However, when comparing local EDPs, e.g., strains and crack widths, significant differences can be obtained. These differences are mainly caused by the torque of the bricks inside the walls (Mann and Müller, 1982). We show that this distortion gets less important with increasing shear span and that for a wall with higher shear span the assumption of a continuous material gives reasonable results even at the local level.