The White Bear Suppression Inventory (WBSI; Wegner & Zanakos, 1994) was originally designed to assess people's inclination toward thought suppression. In this article, we provide a detailed review of previous findings on the structure of this instrument and present a study that took a new statistical approach. It involved an exploratory factor analysis of the French WBSI using the weighted least squares mean and variance estimator as well as parametric item response theory analyses. Results clearly supported a 2-factor structure with a "suppression" and an "intrusion" dimension. Follow-up regression analyses revealed that intrusion significantly predicted anxiety and depression scores, whereas suppression did not.