BACKGROUND: It has been proposed that worry in individuals with Generalized Anxiety Disorder may be reinforced by a positive effect of worry on decision making, as reflected by a steeper learning curve on the Iowa Gambling Task (IGT). We hypothesized that this apparent positive effect of worry is dependent on the IGT parameters, in particular the absence of an opportunity to avoid decisions. OBJECTIVE: (1) To replicate previous findings on the effect of worry on IGT performance. (2) To examine the influence of avoidance opportunity on IGT performance. We hypothesized that the positive effect of worry on learning would be abolished or reversed by the opportunity to avoid. METHOD: A standard IGT and a new IGT version that includes a pass (avoidance) option were completed by 78 and 79 participants, respectively. RESULTS: A beneficial effect of worry on learning in the standard version of the IGT was not observed. In the pass version of the IGT, worry status and avoidance were negatively associated with performance. Worry was not related, however, to pass usage. The hypothesized mediating effect of avoidance was non-significant. LIMITATIONS: It is unclear to what extent these findings generalize to real-life decision making and how clinical status affects results. CONCLUSION: The possibility to avoid a decision results in poorer IGT performance in high relative to low trait worriers. Possible explanations for these findings are discussed.