This paper presents the initial development and validation for a numerical model of a building in Wellington that was instrumented with strong motion recording devices during a number of large earthquakes and aftershocks. The model considers both the structural systems, the underlying soil, and makes consideration for soil structure interaction effects. This validation study aims to demonstrate the predictive capability of the model through comparison with the responses recorded by the instrumented structure and provide a quantitative benchmark for the level of confidence that is reasonable to assign to forward prediction analyses using similar models. This paper discusses ongoing preliminary work towards these goals.