Regrettably, evidence-based practice is often reduced to the provision of empirically-supported treatments, which mostly means CBT. While this may suit government bodies and others who want simple guides to legitimate practice, it has little correspondence with reality. In fact the evidence suggests that the form of therapeutic intervention plays a minor role in client outcome. Some have argued non-specific or common factors such as therapeutic alliance are the key to client outcome but it is far from clear that such factors are under the control of the therapist. A practice-based evidence approach whereby the therapist receives, and makes use of, feedback on how the client is progressing has also been advocated. This paper argues that it is not good enough to simply insist that practitioners adopt evidence-based practice. Rather we must ensure practitioners have a clear and nuanced understanding of the evidence to make appropriate clinical decisions and, where necessary, defend those decisions. The presentation reviews the current state of the evidence and explores how it might be used in clinical decision-making that is consistent with the tenets of evidence-based practice.