Abstract: The fourth industrial revolution has seen more business leverage technology such as artificial intelligence to improve their workflows and gain a competitive advantage in their sector. The aim of this project was to implement a predictive system to predict the selling and asking price of used cars in the South Africa market. This was in order to provide valuable insight for users and thereby decrease the time taken for users to estimate an asking price of a used vehicle. The predictive system consisted of an ensemble of models which was a combination of a one-time linear regression model trained on a returned query set and a decision tree trained on all available data on the auto trader platform. The ensemble approach was combined using a weighting scheme which was based on two measures, the likelihood that the target mileage was part of the distribution and its distance to the returned query results. It was shown that the decision tree had a R squared score of 0.895 which suggest that it fitted well on the available data and thus provides sensible predictions for the cost of used cars in South African market.