The need for psychotherapy research to understand the therapist effect has been emphasized in several studies. In a large naturalistic study (255 patients, 70 therapists), this topic was addressed using therapists' self-assessed difficulties in practice and interpersonal functioning in therapeutic work as predictors of patient outcome in three conventional outcome measures. Three-level growth curve analyses were employed to assess whether the therapist characteristics, measured by the Development of Psychotherapists Common Core Questionnaire (Orlinsky & Rønnestad, 2005), predicted the level of and change in patient symptom distress (SCL-90R), interpersonal problems (IIP-64), and observer-rated global functioning (GAF). Preliminary estimates of therapist effects in patient change indicated that 4% of change in general symptom distress (GSI), almost 21% of change in IIP global scores, and 28% of growth in GAF could be attributed to therapist differences. The results also demonstrated that certain therapist self-perceptions were clearly related to patient outcome. For example, therapists' scores on a type of difficulty in practice called "Professional self-doubt" (PSD) (denoting doubt about one's professional efficacy) were positively associated with change in IIP global scores. It is suggested that therapists' self-reported functioning can be of value in understanding how individual therapists contribute to therapeutic change although their influence is not necessarily exerted in expected directions.