Cognitive models of obsessive-compulsive disorder propose that beliefs about the importance of and need to control thoughts (ICT) are central to the maintenance of the disorder. Cognitive Bias Modification for Interpretation (CBM-I) can be used to experimentally test this theory and may also have clinical utility as an adjunct therapeutic tool. The current study extended previous research to investigate whether two CBM-I sessions (one within and one outside the laboratory) would augment effects on obsessive-compulsive beliefs and behavior. We randomly allocated undergraduate participants high in ICT beliefs to a Positive (n = 30) or Control (n = 36) CBM-I condition and conducted multi-modal assessments immediately following the first training and at one-week follow-up. As predicted, participants in the Positive condition reported a reduction in obsessive-compulsive beliefs from baseline to follow-up (partial η
2 = .42), whereas those in the Control condition did not. Participants responded more adaptively to the ICT relevant stressor task at follow-up compared to post-intervention, but there was no significant difference between conditions. Likewise, participants reported a reduction in obsessive symptoms over time that did not differ between conditions. The findings are considered in light of cognitive models of OCD, and clinical implications are discussed.