Objectives: The aim of the present study was to investigate “Proactive-Adjustment hypothesis” (PA) during the Stop Signal Task (SST). The PA is implied in the highly inconsistent literature, and it deals with the role of response inhibition (RI) in obsessive-compulsive disorder (OCD). This hypothesis assumed that participants would balance stopping and going by adjusting the response threshold (RT) in the go task. We verified whether the PA strategy was also implemented in our clinical group. Methods: To reach this goal, we analyzed SST performances in a group of 36 patients with OCD and 36 healthy controls (HCs). To identify different participants’ behaviors during the task, without preconceived notions regarding the diagnosis, we performed a cluster analysis. Furthermore, we analyzed the influence of drug therapy and we investigated whether the rule and reversal acquisition investigated with the Intra-Extra Dimensional Set Shift, differed in the two clusters. Results: We did not find any difference relative to the number of patients with OCD and HCs included in the two clusters. Furthermore, we found that only Not Proactive participants performed the task as fast as possible, while Proactive participants consistently slowed down their RTs and showed a lower number of Direction Errors, higher Stop Signal Delay, and worse cognitive flexibility. Conclusions: Our results show that among patients with OCD the use of PA is changeable and does not differ from HCs. This finding supports the idea that the RI heterogeneity concerning patients with OCD could be related to PA. (JINS, 2018, 24, 1–12)