Introduction: While ample data demonstrate the effectiveness of inpatient psychosomatic treatment, clinical observation and empirical evidence demonstrate that not all patients benefit equally from established therapeutic methods. Especially patients with a comorbid personality disorder often show reduced therapeutic success compared to other patient groups. Due to the heterogeneous and categorical personality assessment, previous studies indicated no uniform direction of this influence. This complicates the derivation of therapeutic recommendations for mental disorders with comorbid personality pathology. Methods: Analyzing n = 2094 patients from German university hospitals enrolled in the prospective “MEPP” study, we tested the dynamic interaction between dimensionally assessed personality functioning and psychopathology of anxiety and depression. Results: Longitudinal structural equation modelling replicated the finding that the severity of symptoms at admission predicts symptom improvement within the same symptom domain. In addition, we here report a significant coupling parameter between the baseline level of personality function and the change in general psychopathology - and vice versa. Discussion and conclusion: These results imply that personality pathology at admission hinders the therapeutic improvement in anxiety and depression, and that improvement of personality pathology is hindered by general psychopathology. Furthermore, the covariance between both domains supports the assumption that personality functioning and general psychopathology cannot be clearly distinguished and adversely influence each other. A dimensional assessment of the personality pathology is therefore recommendable for psychotherapy research and targeted therapeutic treatment.