Objective The aim of the study was to evaluate decision-making process and probabilistic bias in patients with obsessive-compulsive disorder (OCD). Methods Our sample consisted of 80 subjects (40 diagnosed with OCD and 40 healthy controls). To assess the clinical characteristic of the sample we performed a clinical interview and a psychometric assessement using the Barrat Impulsiveness Scale (BIS-11) and the Cloniger's Temperamnt and Character Inventory (TCI-125). To assess decision making processes and probabilistic reasoning we used the Iowa Gambling Task (IGT) and the Beads Task. Results OCD group had a significantly lower IGT final netscore (p 0.032). Moreover, controls showed a significant improvement between netscore 1 and netscore 5 (p 0.014) while patients did not (p 0.700). Patients also showed significantly less draws till decision on the Beads Task (p 0.000). We performed four regression analysis to evaluate the influence of clinical variables (duration of illness, treatment resistance, Y-BOCS Score, presence of tics, Barratt Score) and personality traits (TCI sub-scales) on IGT netscore and Beads Task. The results indicated that the four overall models were not statistically significant. Conclusions Our data show the presence of a decision making and probabilistic reasoning impairment in OCD. These data could be intareptated in a neurocognitive perspective as executive dysfunction and in a neuroeconomic perspective as a reduced loss avoidance and overconfidence during decision making processes.