An innovative methodology is presented for identifying and assessing change process in psychoanalytic treatments. Using the Psychotherapy Process Q-set (PQS), a panel of experienced psychoanalysts developed a prototype of an ideal psychoanalytic hour. This prototype was then applied to verbatim transcripts of three archived treatment samples: psychoanalyses, long-term analytic therapies, and brief psychodynamic therapies. The degree to which these treatments fostered an analytic process as represented by the prototype was measured quantitatively. Analytic process was significantly more present in psychoanalyses than in the long-term analytic therapies, which, in turn fostered significantly more analytic process than did brief psychodynamic therapies. The study demonstrates that, given descriptive language that does not represent a particular theoretical perspective, analysts can agree on a definition of analytic process, and that analytic process can be operationalized and quantitatively assessed. A second study demonstrates that despite consensus on its definition, there is not just one proper analytic process; rather, there are change processes unique to each dyad. Two quantitative case studies illustrate how each analytic pair has a unique interaction pattern linked to treatment progress. These dyad-unique "interaction structures" are recurrent, mutually influencing patterns of interaction, the experience, recognition, and comprehension of which appear to be a fundamental component of therapeutic action. A bipersonal model is described that attempts to bridge theories of therapeutic action that focus on insight and self-understanding and those that emphasize the patient's experience of the therapist.