This study applies a process of correlation analysis and hedonic based modelling to investigate the influence of housing attributes and date of sale on transaction prices for residential real estate in the Nelson Mandela Metropole area. Availability of reliable data is a challenge in real estate studies in South Africa. This study integrates several publicly available resources for the purposes of constructing a statistical model. A model is constructed based on a forward variable selection routine using selection criteria. The resulting model shows six highly significant variables that can be used for modelling purposes. The resulting model is used to make comparisons in the value of the geographical location.