The criteria for the eleven personality disorders were determined by American psychiatrists and evaluated in field trials with American populations.

Trans-cultural comparison using DSM-III-R terminology has limited itself to axis I disorders, and this is understandable: considering the significant differences in evaluation of behavioral characteristics and personality traits by American and by non-American people, it is almost impossible to compare evaluations made by non- American evaluators with the American ones. Behavior and personality traits that are considered pathological in the USA may be considered normal, adaptive and even praiseworthy in another culture and would then not identify a personality disorder. In a pilot study with Hindu-psychiatrists, many of the criteria, especially for cluster C personality disorders, turned out to describe behavior which is considered quite normal and adaptive in a Hindu-society. The criteria for personality disorders will have to be evaluated for each well-defined non-American culture, before DSM-III- R axis II can be applied to its population. Until then, all publications regarding the epidemiology of personality disorders in `the normal population' will only be of anecdotal value.