The Yale-Brown Obsessive Compulsive Scale (Y-BOCS) and different versions of the Padua Inventory (PI) are frequently used instruments to measure symptoms of obsessive-compulsive disorder (OCD). However, little is known of how these different versions of the PI compare to each other in their sensitivity to measuring treatment outcome, and there is currently no adequate explanation to account for the weak relationships between self-report measures and the Y-BOCS. This study aimed to investigate the sensitivity of these measures to treatment outcome, and to examine whether differences in how they measure symptom severity can explain the weak relationships. Hypotheses were: (1) the Y-BOCS would be significantly more sensitive to measuring treatment outcome than the PI versions; (2) correlations between the measures would be significantly stronger for change scores as compared to relations measured at a single point in time; (3) weak relationships can be explained by the PI measuring symptom severity based on content and the Y-BOCS measuring symptoms, independent of content. Results showed that the Y-BOCS was significantly more sensitive to measuring treatment outcome than the PI versions, while differences between the questionnaires in which severity is measured can provide a partial account for why weak relations are observed between these measures.