In 2004, the Bangor Gambling Task (BGT) was put forward as a simplified of the Iowa Gambling Task (IGT) - one of the primary experimental paradigms used to assess emotion-based decision making under uncertainty. We aimed to investigate the suitability of the BGT as a longer-term re-assessment tool of the IGT. Method - Using a N = 176 focused subsample of the large ongoing prospective cohort study Tracking Adolescents’ Individual Lives Survey (TRAILS), we investigated three suitability criteria with regard to 1) incremental learning patterns within each task, 2) correlation between the net-scores of tasks, and 3) within-level change in emotion-based decision-making by social stress as predicted by evolutionary theories. One out of three suitability criteria were met, that is, we found similar incremental learning patterns in gambling performance as reported in 2004, but no correlation between net scores of both tasks, nor a decrease in approach-related gambling choices in individuals exposed to social adversity versus controls. Although the BGT may be a useful tool for clinical neuropsychologists, it may be less suited to test changes in adolescents’ decision making performance in combination with the IGT over a larger time frame.