This paper discusses the application of Bayesian networks for evaluating evidence in forensic fire investigations, building upon a previous article in the series. It presents two casework examples to demonstrate the practical use of these networks.