In today?s increasingly electronic society and with the rapid advances of electronic commerce on the Internet, the use of credit card transactions have become the standard for Internet and Web-based e-commerce. In this paper, we present a credit card fraud detection model which takes into account present as well as the past behaviour. The detection model comprises of card validation via Luhn?s algorithm, two initial probability assignments based on address mismatch and spending pattern, an advanced combination heuristic, spending-pattern database and Bayes? theorem. Advanced combination heuristic is an improvement that eliminates the conflict of existing Dempster-Shafer theory.