Collusion is the deliberate cooperation of two or more parties to the detriment of others. While this behaviour can be highly profitable for colluders (for example, in auctions and online games), it is considered illegal and unfair in many sequential decision-making domains and presents many challenging problems in these systems. In this thesis we present an automatic collusion detection method for extensive form games. This method uses a novel object, called a collusion table, that aims to capture the effects of collusive behaviour on the utility of players without committing to any particular pattern of behaviour. We also introduce a general method for developing collusive agents which was necessary to create a dataset of labelled colluding and normal agents. The effectiveness of our collusion detection method is demonstrated experimentally. Our results show that this method provides promising accuracy, detecting collusion by both strong and weak agents.