OBJECTIVE: The authors developed a simplified method of rating the severity of personality disorder. METHOD: The new rating method is based on four levels of severity: no personality disorder, personality difficulty, simple personality disorder, and diffuse personality disorder. The new method was applied to different diagnostic systems and was then compared with an old rating system based on six severity levels. Data were derived from a longitudinal study in which 163 patients with anxiety and depressive disorders had initial assessments of personality status and were followed up over 2 years. Ratings of psychiatric symptoms were made by using the Comprehensive Psychopathological Rating Scale over this period. The results were analyzed with special attention to linear and quadratic trends. RESULTS: The new system was clinically useful in separating patients' initial assessments and outcomes. Patients with no personality disorder had the lowest initial symptom scores and the best outcomes, and those with diffuse personality disorder had the highest initial levels of symptoms and improved least over the 2 years. When the patients were separated by the old classification system, 72% of the variation between groups was accounted for by linear and quadratic trends; the comparable percentage was 97% when the patients were categorized by the new system. CONCLUSIONS: The new system of rating severity of personality disturbance is an improvement on existing methods and allows ratings to be made easily from DSM-IV and ICD-10.