A seminal hypothesis underlying compulsive behaviour in obsessive compulsive disorder (OCD) is diminished goal-directed control, resulting in the reliance of the habit system and thus the performance of ridgid, repetitive acts. This beahvioural control issue has further been demostrated to be well captured and specific to a transdiagnostic dimension of compulsivity, in opposition to other dimensions of psychopathology like anxious-depression and social withdrawal. In this talk, I will present some research suggesting that OCD is linked to dsysfunctions beyond disruption in action systems - instead, a dyfunction in their mental models that action relies upon. I will discuss several metacongitive deficits associated with compulsivity which were not previously revealed in case-control designs, which suggest that mental representations may be awry. I will then show that these diminished mental representations could underlie the poor goal-directed control notable in compulsivity. Finally, I will talk about some preliminary work on investigating biased thought patterns as an avenue to probe mental representations that underlie obsessionality, the other aspect of OCD.