Worry and rumination are two forms of repetitive thinking characterised by their negative content and apparently uncontrollable nature. Although worry and rumination share common features, it remains unclear whether they are both maintained by the same cognitive mechanisms. In this talk, I will present research investigating the tendency to generate negative interpretations in regards to ambiguous information (termed negative interpretation bias) as part of cognitive mechanism underlying repetitive negative thinking. To assess the causal role of negative interpretation bias, I will present two studies in which we compared a multi-session training program designed to promote more positive interpretations to a control condition to examine whether repeatedly training positive interpretations would reduce worry and rumination in individuals with generalised anxiety disorder or depression (Study 1) and high levels of worry or rumination (Study 2). In both studies, participants received a three-week internet-based intervention consisting of one initial session at the first study visit and 9 home-based sessions of training (or active control). Effects of the training on measures of interpretation bias (closely related to the training) as well as outcomes related to repetitive negative thinking and psychological distress were examined, and additionally, we investigated the impact of our training vs active control on questionnaire measures at 1-month follow-up. I will discuss our findings in light of shared cognitive mechanisms contributing to the maintenance of worry and rumination in generalised anxiety disorder and depression.