Background: The Freiburger Personality Inventory (FPI-R) is a well established and proven instrument for the assessment of personality traits. Although personality is conceived as a stable trait, clinical experience indicates that impressive changes are found on personality scales during intensive treatment. Method A large sample of inpatients which were treated with cognitive-behavioral therapy for bulimia nervosa, tinnitus or anxiety disorder was evaluated concerning the question which items of the FPI-R were answered differently or identically before and after intensive therapy. Results: It could be found that items which cover aspects that are central to the therapy more often show changing answers. The use of conditional form and indefinite frequency adjuncts in the formulation of items evidently allowed a more differentiated weighting of pros and cons at the end of therapy. Effects of regression to the mean could be excluded as an explanation by empirical data. Conclusion: It can be concluded that changes in answering items before and after intensive therapy can be explained as specific effects of therapy.