A conceptual and empirical critique of Kernberg's influential object relations theory is presented as a case study of the limitations of structural ontological presuppositions in accounting for psychological processes. A summary overview is provided of Kernberg's systems model, the process of internalization, his developmental stages, and his conception of the borderline personality organization. Then a detailed critique considers: (1) the foundations of Kernberg's model in affective memory and units of internalized object relations; (2) the principles of construction underlying his developmental processes; (3) the resulting product of these processes, his structural model and theory of the borderline personality organization; and (4) the relationship between his metapsychology and his clinical theory. Suggestions are made for how process models of personality and psychopathology can redress the problems with structural accounts.