Cognitive biases are systematic errors that result from reliance on heuristics in decision-making. Such biases are typically automatic and unconscious influences on behavior, and can occur in a wide range of situations and contexts. Cognitive biases are generally resistant to mitigation training. This project adopted a novel approach to develop computer game-based training to attempt to mitigate three forms of cognitive bias: fundamental attribution error, the tendency to assume dispositional rather than situational influences account for behavior of others; confirmation bias, the tendency to seek and remember information that matches or supports one's view; and bias blind spot, the tendency to regard one's own decisions as being free from cognitive bias, even where one can recognize that bias in others. Participants were randomly assigned to play the training game once, or repeated twice with a 7-10 day delay between sessions (mean duration first play=43 minutes; second play=34 minutes), or to a control condition that employed a 30- minute professionally developed training video. Effects of training were measured on external questionnaire-based items, both immediately post-exposure, and at an 8-week retention interval. The game was intended to develop conceptual understanding of these biases, and recognition of circumstances within which they might occur. Using notional “toolsâ€� presented within the game, participants learned and practiced strategies to avoid decision-making influenced by the cognitive biases. Results showed that the training game successfully reduced bias on the assessment instrument, and outperformed the video both immediately post-training and at the retention test. Repetition of the training game did not further advantage immediate post-test performance but significantly improved retention. Validation of the key findings was confirmed by an independent group who used the training game with their own novel bias assessment instruments (to which the researchers and game-developers had no access or content information).