In the process of developing a two-part therapist rating scale, we obtained questionnaire responses from a large sample of subjects who had undergone one of a variety of psychotherapies. Our data support the importance of perceived therapist care as a strong correlate of patient satisfaction, and patients seemed able to discriminate clearly between psychodynamic and cognitive-behavioral therapies. In the first (nonspecific) questionnaire, we derived five dimensions (Care-Concern, Directive-Control, Critical-Confronting, Understanding, and Charisma) from principal-components analyses, with the first two dimensions being relatively independent of each other. A multiple-regression analysis revealed that combined scale scores accounted for 50% of the variance of patient satisfaction ratings, with the Care dimension accounting for most of this effect. In the second questionnaire, data were analyzed from respondents who had undergone therapy with known cognitive-behavioral or psychodynamic therapists; but, in a minority of subjects who were recruited through the media, we depended on detailed descriptive vignettes to allocate them to one or another of these "specific" types of psychotherapy. Six scales emerged, three characterizing cognitive-behavioral therapies (Behavioral Task Orientation, Organization, and Cognitive Focus) and three representing psychodynamic therapies (Transference, Inner Conflict, and Relationship). Subjects' ratings on all scales strongly discriminated between the two broad modalities of therapy, suggesting that patient reports may provide valid measures of therapist styles. We conclude that questionnaires such as the one we are developing may provide economical measures for assessing therapist consistency when comparing different modalities of psychotherapy.