The Evidence Base for Relational and Psychodynamic Therapy: What the Research Actually Shows
If you trained in a CBT-dominant environment, you've probably absorbed the message — sometimes explicitly, sometimes just by implication — that psychodynamic and relational approaches are less evidence-based than structured, protocol-driven therapies.
That message is not accurate. And for psychologists considering whether to deepen their practice in this direction, it's worth spending some time with what the research actually shows.
This page summarises the key findings in plain language. It's written for practising clinicians, not academics — so rather than exhaustive citation lists, you'll find the main conclusions, what they mean for clinical practice, and where to look if you want to go deeper.
The Short Answer
Psychodynamic and relational therapy is evidence-based. Across dozens of meta-analyses and hundreds of randomised controlled trials, it has demonstrated effectiveness for a wide range of common mental health presentations — comparable in effect size to cognitive-behavioural therapies, and with outcomes that hold up at follow-up in a way that distinguishes it from many shorter-term approaches.
The perception that it lacks empirical support is a product of historical imbalance in research funding and dissemination, not a reflection of the actual evidence.
What the Key Research Shows
Shedler (2010) — American Psychologist
The most widely cited single paper on this question, published in the American Psychological Association's flagship journal, found that effect sizes for psychodynamic therapy are as large as those reported for therapies actively promoted as empirically supported — including CBT. Across studies of short-term psychodynamic therapy, the overall effect size for symptom improvement was 0.97. At long-term follow-up of nine months or more post-treatment, effect sizes increased to 1.51, suggesting patients continue to improve after treatment ends — a finding that is not typically seen with structured, protocol-based approaches.
Shedler also noted that many non-psychodynamic therapies may be effective in part because skilled practitioners incorporate techniques that have long been central to psychodynamic practice: attention to the therapeutic relationship, exploration of emotion, and work with avoidance and resistance.
Leichsenring et al. (2023) — World Psychiatry
This umbrella review — one of the most comprehensive assessments of the evidence to date — evaluated psychodynamic therapy as an empirically supported treatment across common mental disorders in adults. Published in World Psychiatry, one of the highest-impact journals in the field, it found moderate-quality evidence for psychodynamic therapy's effectiveness across depressive disorders, anxiety disorders, somatic symptom disorders, and personality disorders. Critically, it found the evidence comparable to other active treatments — meaning the historical framing of CBT as the gold standard and psychodynamic approaches as unproven does not reflect the current state of the literature.
Leichsenring and Leibing (2003) — American Journal of Psychiatry
A meta-analysis of psychodynamic therapy and CBT for personality disorders found a large overall effect size for psychodynamic therapy (1.46), including effects sustained at follow-up. Personality disorders are among the most clinically demanding presentations any psychologist will encounter — and the evidence here for psychodynamic approaches is robust.
Why Outcomes Improve After Treatment Ends
One of the most clinically significant findings in psychodynamic research is the sleeper effect — the tendency for outcomes to continue improving after therapy concludes.
This is not common across all therapeutic modalities. It suggests that psychodynamic work produces something different from symptom reduction: an increase in the client's capacity to manage their own internal world, understand their relational patterns, and make meaning of their experience.
Shedler's 2010 meta-analysis documented this pattern systematically. More recent research continues to replicate it. For the clinician, this has direct implications: the goal is not just to resolve the presenting problem but to build psychological resources that keep working when the therapy is over.
This is also what makes psychodynamic work intrinsically meaningful for many practitioners. The change you're facilitating is deeper and more durable than symptom management.
The Neuroscience and Attachment Dimension
The evidence base for relational and psychodynamic approaches extends beyond clinical outcome research into neuroscience — specifically, what decades of attachment research and interpersonal neurobiology tell us about how therapeutic change actually occurs.
The work of Peter Fonagy and his colleagues at University College London has been particularly influential. Fonagy's research on mentalisation — the capacity to understand one's own and others' mental states — demonstrates that this capacity is central to psychological health, and that it develops primarily within relational contexts. When it's compromised (as it often is in clients with complex presentations), the therapeutic relationship itself is the primary mechanism through which it can be rebuilt.
This gives the relational dimension of therapy not just theoretical weight, but neurological grounding. The experience of being genuinely understood by another person — having one's internal states recognised and reflected back accurately — activates the same neurological processes that support secure attachment. Mentalisation-based therapy, grounded in this research, now has substantial evidence for personality disorders, chronic depression, and complex trauma.
For the practising psychologist, the implication is straightforward: the relationship is not just a vehicle for technique delivery. In many presentations, it is the primary driver of change. Developing the skills to use that relationship deliberately and skillfully is what psychodynamic and relational training is fundamentally about.
What the Evidence Covers and Where the Gaps Are
Being honest about the evidence means acknowledging what it doesn't show, as well as what it does. Most randomised controlled trials of psychodynamic therapy have focused on shorter-term, manualised versions of the approach. Long-term open-ended psychodynamic work — which is closer to what many clinicians actually practise — has a smaller RCT base, partly because it is harder to study with standard trial designs and partly because of historical research funding priorities.
This doesn't mean long-term psychodynamic work is ineffective. Naturalistic studies and effectiveness research consistently show positive outcomes. But psychologists trained to evaluate evidence hierarchies should be aware that the RCT evidence base is stronger for short-to-medium-term structured psychodynamic approaches than for long-term open work.
The evidence base is also stronger for some presentations than others. It is particularly robust for depression, anxiety disorders, somatic presentations, and personality disorders. It is more limited for psychosis and conditions where biological factors are primary drivers.
What the evidence clearly does not support is the claim that psychodynamic therapy is unproven. That framing belongs to an earlier period of the field's history and has not kept pace with the research.
What This Means for Your Practice
For psychologists trained primarily in CBT or structured approaches, this evidence matters in two ways. First, it provides a grounding for expanding your clinical repertoire without abandoning your commitment to evidence-based practice. Adding psychodynamic and relational skills is not a departure from evidence-based care — it is an extension of it, into areas where the evidence is strong and where structured approaches have well-documented limitations.
Second, it changes how you might think about the harder clinical problems. When a client isn't responding to protocol, the research suggests the relational and unconscious dimensions of what's happening in the room are not peripheral — they may be central. Learning to work with those dimensions is not a soft alternative to evidence-based practice. It is itself evidence-based.
If psychologist burnout is part of what's driving your interest in this — the exhaustion that comes from working without a framework adequate to the complexity of what your clients bring — you can read more about that on the psychologist burnout page.
And if you're thinking about how psychodynamic training fits into your CPD requirements as a registered psychologist, that's covered on the CPD for psychologists page.
Frequently Asked Questions
Is psychodynamic therapy as effective as CBT? The current research suggests comparable effectiveness for most common presentations. Shedler's 2010 meta-analysis in the American Psychologist found effect sizes for psychodynamic therapy equal to those for treatments actively promoted as empirically supported, including CBT. Leichsenring's 2023 umbrella review in World Psychiatry reached similar conclusions. Neither approach is uniformly superior — the fit between client, presentation, and modality matters more than which therapy wins a comparison study.
Does psychodynamic therapy work for complex presentations like personality disorders? Yes — and the evidence here is particularly strong. A 2003 meta-analysis by Leichsenring and Leibing found a large overall effect size for psychodynamic therapy with personality disorders. Mentalisation-based therapy (Fonagy and Bateman) has strong RCT support for borderline personality disorder specifically. These are among the presentations where structured symptom-focused approaches most commonly run out of runway.
Are the gains from psychodynamic therapy lasting? The sleeper effect documented in psychodynamic research — where outcomes continue to improve after therapy ends — is one of the most distinctive features of this modality's evidence base. Follow-up data consistently shows maintained and often improved outcomes at nine months to several years post-treatment, which is not reliably demonstrated by shorter-term symptom-focused approaches.
Can I integrate psychodynamic approaches with my existing CBT practice? Yes, and many psychologists do. The research actually suggests that skilled CBT practitioners often already use techniques central to psychodynamic practice — exploration of emotion, attention to the therapeutic relationship, work with avoidance — without necessarily naming them as such. Formalising that knowledge through psychodynamic training gives you a more coherent framework and extends your range with complex presentations.
Ready to Work Differently?
The evidence is clear enough. The more useful question for most clinicians isn't whether psychodynamic and relational therapy works — it's whether you have the training to use it well.
Deep Mind Psychodynamic Training was built for practising psychologists who want to develop that depth. Not as a full retraining, but as a structured, practical deepening of what you already do.
Or if you'd like to experience the approach before committing, join the free June webinar: The Masterful Art of Socratic Questioning
Tania Kalkidis is a registered clinical psychologist (AHPRA PSY0000976980), member of the Australian Psychological Society and Australian Association of Psychologists Inc, and founder of Deep Mind Psychodynamic Training.
Find out now how your own unique attachment style shows up in your work.
