Dissatisfaction with the existing categorical classification system for personality disorders has stimulated DSM-V Research Planning Work Groups to check alternative dimensional representations. Four different strategies have been suggested: 1. dimensional representations of existing categories; 2. dimensional reorganization of diagnostic criteria; 3. integration of personality disorders with dimensional models of general personality structure; 4. identification of spectra of dysfunction cutting across personality, axis I and axis II disorders. Together with models of dimensional classification two further relevant aspects are discussed: First, the question of stability and changeability of personality (disorder) and thus the question of childhood and adolescence antecedents. Second, the search for neuroscientific foundations of personality disorder traits that can be integrated by an endophenotypic approach. These considerations imply a heuristic framework guided by the vision of an etiologically based classification system for personality disorders (and possibly other psychiatric syndromes).