Objective: This study aimed to analyze the characteristics and reasons of early discontinuation of obsessive-compulsive disorder (OCD) trials registered on ClinicalTrials.gov.Methods: OCD trials and relevant publications were searched on ClinicalTrials.gov and PubMed, respectively. The characteristics and details regarding the timely publication of trials were recorded. Cox regression analysis was used to explore factors associated with the early discontinuation of OCD trials.Results: The analysis included 298 OCD therapy trials. Most investigations recruited <100 patients and were more likely to involve adults. Of all OCD studies identified, 67.8% were randomized and 61.4% were blind (single- or double-blind). Universities and hospitals were recorded as the two primary locations in the majority of trials. A total of 155 trials (52%) were completed; however, only 29% of those were published. Of the published trials, >70% were published at least 1 year after completion. Behavioral therapy trials were the most common type of major treatment-aimed OCD trials (39%), followed by drug trials (35.1%) and device/procedure trials (24.7%). The univariate Cox regression analysis indicated that drug trials [hazard ratio (HR) = 2.56, 95% confidence interval (CI): 1.21–5.43], absence of collaborators (HR = 3.87, 95% CI: 1.62–9.26), and sponsorship by industry (HR = 3.97, 95% CI: 1.49–10.53) were risk factors for early discontinuation of OCD trials. Further multivariate Cox regression showed that drug trials (HR = 3.93, 95% CI: 1.71–9.08) and absence of collaborators (HR = 5.17, 95% CI: 1.97–13.54) were independent risk factors for early trial discontinuation of OCD trials. The sensitivity analysis confirmed these results. Non-drug trials (OR = 3.32, 95% CI: 1.21–9.11), absence of collaborators (OR = 3.25, 95% CI: 1.10–9.60), and non-blinded trials (OR = 5.23, 95% CI: 1.05–26.2) were independent risk factors for unreported results in registry.Conclusion: The diagnosis and prevention of OCD are rarely investigated in trials. Underreporting and delayed reporting remain major problems. The type of intervention and participation of collaborators are associated with early discontinuation of OCD trials.