Introduction: Obsessive compulsive disorder (OCD) is a chronic psychiatric disorder where most patients do not reach full symptomatic remission. Identifying predictors of course can improve patients’ care by informing clinicians on prognosis and enhancing treatment strategies. Several predictors associated with improved outcome of OCD were identified. However, research focused mainly on clinical, illness-related predictors of the course of OCD. This study examined the contribution of environmental and interpersonal predictors on the long-term outcome of OCD, in addition to the previously identified clinical indicators. Methods: We used the baseline, two and four-year data of 382 adult OCD patients participating in the naturalistic cohort study of the Netherlands Obsessive Compulsive Disorder Association (NOCDA). Remission was assessed using the Yale-Brown Obsessive Compulsive Scale (Y-BOCS). Predictors of outcome were assessed at baseline, via clinician-rated and self-report instruments. Results: Remission at two and at four-year follow-up ranged from 11% to 26%. Early age of onset and the presence of childhood trauma predicted a worse four-year course. Secure attachment style emerged as a protective predictor of improved outcome. Limitations: The naturalistic design of our study did not enable a systematic estimation the effect of treatments received during the follow-up period. Furthermore, age of onset and childhood trauma were assessed retrospectively, which may contribute to recall bias. Conclusion: Results coincide with previous prediction research and stress the importance of adaptive interpersonal functioning in the course of OCD. Clinical implications and future research directions are discussed.