Jurisdictions in Canada, notably in Ontario and Quebec, are proposing a laudable goal of increasing publicly funded access to psychotherapy. Ontario and Quebec will likely follow the lead of the Increasing Access to Psychotherapy (IAPT) program in the United Kingdom and train psychotherapists to provide cognitive–behavioural therapy. Results from IAPT provide some important lessons about taking an approach that prefers one brand of psychotherapy. We argue that such policy decisions are based on a medical model approach to psychotherapy that makes erroneous assumptions about what psychotherapy is, what is the nature of the evidence, and how training and services should be delivered. In this article, we review critically these assumptions and the state of the art of the research about what makes psychotherapy work. Psychotherapy is effective and preferable to antidepressant medication. Differences between psychotherapies account for a small proportion of variance in patient outcomes. The largest known predictors of patient outcomes are patient factors (coping style, resistance/reactance, interpersonal problems, culture) and therapeutic relationship factors (therapeutic alliance). Also notable are therapist factors (facilitative interpersonal skills, empathy, and managing countertransference), and practices like progress monitoring. Canadian jurisdictions should note that training therapists to adapt treatment and interpersonal stances to patient characteristics, to develop and maintain the therapeutic relationship, to enhance therapist facilitative interpersonal skills, and to engage in progress monitoring has a greater chance of achieving a reduction in the burden caused by depression and anxiety among their citizens than focusing on delivering a particular brand of psychotherapy. (PsycINFO Database Record (c) 2018 APA, all rights reserved)