Four therapists' participation in six therapies were analyzed at the utterance level. Therapist's utterances ( N = 5,504) were rated on 39 categories of the Stuttgart Interactional Category System in two early, two middle, and two late sessions, providing a primary and a cross-validation sample. Application of the P-technique revealed four stable and reliable therapist participatory factors (i.e., Objective Information Exchange Seeking, Directing InsightfuUPainful Work, Self-Involving Disclosure, and Advice Giving Information), accounting for 39% of the total variance. Multivariate analyses revealed a signrfcant therapist by phase interaction on all factors, with most differences occurring on the work and disclosure factors in the middle phase and between two groups of therapists (within treatments). Further examination using chronographic analyses revealed ( 1 ) some shortcomings in the analysis of variance approach, and (2) significant strengths in identrfying naturally occurring episodes of therapist participation. Results are discussed in terms of the need to further develop the use of the P-technique to discover generalizable structures of therapist participation and to track quantitative variations in such structures using chronographic analysis.Psychotherapy researchers typically ask if treatment is effective, and what parameters of participants and treatment promote success. Consequently, studies of effectiveness have taken precedence over studies of therapeutic process (Omer & Dar, 1992). The field simply places a lower value on investigations that are not focused on outcome. The low value placed on process studies seems to derive from defining investigations of therapy as primarily a study of treatment: To fulfill their scientific mission such studies must address the practical issues of cure, accountability, and cost-effectiveness.