The theoretical framework developed considers the objectives of the colluders and the institutional framework in which they operate. Aspects of the bidding behavior of collusive groups which differentiate them from contractors entering competitive bids were identified, and empirical analogs of these aspects of behavior were developed as indicators of collusive bidding. Using a sample of bid situations, models were then calibrated to determine the practical importance of these indicators. Finally, the models were used to forecast which of an independent sample of bid situations involved collusion. The theory of collusive behavior is that a cartel of price-fixers is usually pursuing the objective of raising the profits of its members above the normal level. As part of this effort, they must also monitor individual cartel members and avoid detection by outside agencies responsible for enforcing antitrust laws. Since it must simultaneously monitor, avoid detection, and enhance profits, the cartel is bound to take some actions that will reveal its existence. The proposed models provide mathematical techniques for identifying various cartel actions. While the proposed method focuses on identifying contracts which are collusive, it is suggested that by reorganizing the data to yield the bidding history of each contractor, with special attention to the estimated probabilities of collusion by contract included in the models, classification schemes for contractors can be developed. Data from model testing are provided.