CBT vs psychoanalysis: which therapy actually works for anxiety?

Most people choosing therapy ask the wrong question. It is not which therapy is better. It is which therapy is better for what.

Both cognitive behavioral therapy (CBT) and psychodynamic therapy have solid research behind them. Both reduce anxiety more than no treatment at all. The decision depends on the nature of your anxiety, your history, and what you are actually trying to change.

What each therapy actually does

CBT targets the relationship between your thoughts, feelings, and behaviors. Aaron Beck developed the approach in the 1960s and 1970s while treating depression; it was adapted for anxiety disorders through the 1980s and 1990s (Beck, 1976). The core premise: anxiety is maintained by distorted thinking patterns and avoidance behaviors. Identify those patterns, test them against reality, change your behavioral responses, and the anxiety loses its grip.

Sessions are structured. You get homework between appointments. A CBT therapist for panic disorder might help you track automatic thoughts during an attack, challenge catastrophic interpretations ("I am dying" becomes "I am uncomfortable"), and use exposure exercises to face feared situations gradually. The therapy runs 12 to 20 sessions.

Psychodynamic therapy is the modern, shorter descendant of classical psychoanalysis. Freud's original framework (Freud, 1917/1966) involved years of frequent sessions exploring the unconscious through free association and dream interpretation. Contemporary psychodynamic therapy typically runs 16 to 50 sessions. The focus shifts from symptoms to underlying patterns: how early relationships shaped the way you relate to yourself and others, what feelings you avoid without realizing it, how past experience drives present distress.

A psychodynamic therapist will not give you worksheets. The work happens in the therapeutic relationship itself. You talk. The therapist listens for recurring themes, emotional avoidance, and the defense mechanisms that protect you from anxiety at the cost of self-awareness. The goal is insight that changes something at a deeper level than behavior.

What the research says

The evidence base for both therapies is stronger than most people assume, and more comparable than the headlines suggest.

A 2016 meta-analysis by Cuijpers, Cristea, Karyotaki, Reijnders, and Huibers, published in World Psychiatry, examined 144 randomized controlled trials (184 comparisons) of CBT for major depression and anxiety disorders. For anxiety disorders specifically, CBT produced substantial effect sizes, making it one of the most empirically supported treatments available. NICE (National Institute for Health and Care Excellence) recommends CBT as a high-intensity psychological treatment for anxiety disorders under its stepped-care model, with CBT sitting at the top of the care pathway for conditions like GAD, panic disorder, and PTSD (NICE CG113, NG116).

Psychodynamic therapy has historically been at a disadvantage in research terms, largely because it is harder to manualize and therefore attracts fewer industry-funded RCTs. Jonathan Shedler addressed this directly in a landmark 2010 paper published in American Psychologist. Shedler conducted a meta-analysis and found that psychodynamic therapy produces effect sizes of 0.97, which are comparable to those reported for CBT. His conclusion was pointed: the perception that psychodynamic therapy lacks evidence reflects the research funding landscape more than the clinical reality. Patients improve at similar rates.

A naturalistic follow-up by Leichsenring, Salzer, Beutel, and colleagues, published in the American Journal of Psychiatry in 2014, tracked patients with social anxiety disorder from a 2013 RCT through two years after treatment ended. Both CBT and psychodynamic therapy produced significant reductions in anxiety, and both held at the two-year mark. The difference was in trajectory: psychodynamic therapy patients showed continued improvement after sessions ended, while CBT gains stabilised at end-of-treatment levels.

For complex disorders, the evidence tilts toward longer-term psychodynamic work. A 2008 meta-analysis by Leichsenring and Rabung in JAMA reviewed 23 studies and found that long-term psychodynamic psychotherapy outperformed short-term approaches for patients with multiple, chronic, or complex conditions.

Where CBT has the edge

CBT is the stronger choice for specific, well-defined anxiety presentations. The evidence for specific phobias, panic disorder, OCD, and PTSD is particularly robust, with dozens of high-quality RCTs demonstrating consistent results (Cuijpers et al., 2016).

The structured format has practical advantages too. CBT is easier to deliver consistently, easier to train therapists in, easier to replicate across settings. That is partly why it dominates clinical guidelines and insurance coverage: not necessarily because it is more effective, but because it is more measurable and more scalable.

If your anxiety is recent, situational, or tied to a specific trigger, CBT tends to deliver results faster. You can reasonably expect meaningful symptom reduction within 12 to 16 sessions. For someone facing a specific fear, managing panic attacks, or working through PTSD from a defined event, CBT is usually the more direct path to mental clarity.

CBT also suits people who want a skills-based approach: homework between sessions, clear milestones, measurable progress.

Where psychodynamic therapy has the edge

Psychodynamic therapy performs better when the anxiety is long-standing, diffuse, or entangled with how you relate to other people. Generalized anxiety that has been present since childhood, social anxiety rooted in early attachment patterns, or anxiety that seems to migrate from one worry to the next regardless of what changes in your life: these presentations point toward something deeper than a set of maladaptive thoughts.

The Leichsenring et al. (2008) meta-analysis in JAMA is relevant here. When disorders are complex, chronic, or involve personality-level patterns, short-term approaches have higher relapse rates and lower overall effectiveness. Psychodynamic therapy, by addressing the underlying structure rather than just the surface symptoms, produces more durable change in these cases.

Psychodynamic therapy also tends to produce broader changes beyond the presenting symptom. Patients often report improvements in relationships, self-understanding, and the capacity to tolerate difficult emotions, changes that CBT, with its symptom focus, does not always target. Building a growth mindset in the context of chronic anxiety often requires more than behavioral tools. It requires understanding where the anxiety came from.

The honest limitation here is that most head-to-head RCTs run only 12 to 20 sessions, which structurally advantages CBT's shorter, more defined arc. Psychodynamic therapy may need longer to show its full effects, which makes it harder to study in standard trial formats. The research is improving, but the comparison is not yet on perfectly equal footing.

The sleeper effect: why psychodynamic gains often grow after treatment ends

One of the most counterintuitive findings in psychotherapy research comes from Shedler's 2010 analysis in American Psychologist. Psychodynamic therapy patients often continue to improve after treatment ends. This phenomenon, known in psychotherapy research as the "sleeper effect," is a consistent feature of longer-term insight-oriented work.

The mechanism makes sense in light of what the therapy targets. CBT teaches skills that are useful when applied. Psychodynamic therapy aims to change something more fundamental: how you process experience, relate to others, and relate to yourself. Those changes, once set in motion, continue operating after the formal sessions end.

The Leichsenring et al. (2014) social anxiety RCT demonstrated exactly this pattern. At the end of treatment, CBT and psychodynamic therapy showed comparable results. At the two-year follow-up, the psychodynamic group had continued to gain ground. CBT's results had held steady, but the gap had closed, with no statistically significant difference between the two by the end of follow-up.

This is not an argument that psychodynamic therapy is superior overall. It is an argument that the timeframe of measurement matters. Studies that measure outcomes only at the end of a short treatment window may systematically underestimate psychodynamic therapy's effectiveness.

How to choose

Both therapies work. The question is fit.

Choose CBT if your anxiety is tied to a specific trigger or event (phobias, panic attacks, OCD, PTSD from a defined incident), if it has been present for a shorter period with no clear relational roots, or if you want structured sessions, homework, and measurable milestones. It is also the practical choice when you have a limited number of sessions and want maximum symptom reduction within that window.

Choose psychodynamic therapy if your anxiety is longstanding or generalized, if it has persisted through symptom-focused treatment, if it shows up mainly in relationships, or if you have a sense it is connected to your history in ways you have not fully worked out. The tradeoff is time: the approach asks more of it, and you should expect to invest in a longer process.

When the choice is unclear, book consultations with therapists from each tradition and treat them as interviews, not commitments. Starting with one therapy does not lock you in. Some practitioners are trained in both and can integrate elements. Combination approaches exist and show promise, though the evidence base for combined formats is still developing. A skilled therapist you actually connect with will outperform a technically better match you cannot engage with.

---

FAQ

Is CBT or psychodynamic therapy better for generalized anxiety disorder (GAD)?

Both have evidence for GAD. CBT has more RCTs specifically targeting GAD and is the NICE-recommended first-line treatment. Psychodynamic therapy is worth considering when GAD has been chronic, has not responded to CBT, or seems rooted in relational and attachment patterns. Leichsenring et al. (2008) found long-term psychodynamic therapy superior for complex presentations, which includes persistent GAD.

How long does each therapy take?

CBT for anxiety disorders typically runs 12 to 20 sessions. Modern psychodynamic therapy runs 16 to 50 sessions, though long-term versions extend beyond that. Classical psychoanalysis (multiple sessions per week over years) is rarely the default today. Most practitioners offering "psychodynamic therapy" are working in the shorter-term format.

Does therapy work better than medication for anxiety?

Both CBT and psychodynamic therapy produce lasting change after treatment ends, which medication alone generally does not. For moderate-to-severe anxiety disorders, combined treatment (therapy plus medication) often outperforms either alone. The choice depends on the specific disorder, severity, and individual factors. A psychiatrist or GP can advise on whether medication is appropriate alongside therapy.

---

For a deeper look at the unconscious patterns that psychodynamic therapy targets, the article on defense mechanisms covers ten of the most common ones with real-world examples and research backing.

Key sources and further reading

Beck AT. Cognitive Therapy and the Emotional Disorders. International Universities Press, 1976.

Cuijpers P, Cristea IA, Karyotaki E, Reijnders M, Huibers MJH. "How effective are cognitive behavior therapies for major depression and anxiety disorders? A meta-analytic update of the evidence." World Psychiatry. 2016;15(3):245–258.

Freud S. Introductory Lectures on Psychoanalysis. 1917. Norton edition, 1966.

Leichsenring F, Rabung S. "Effectiveness of long-term psychodynamic psychotherapy: A meta-analysis." JAMA. 2008;300(13):1551–1565.

Leichsenring F, Salzer S, Beutel ME, et al. "Long-term outcome of psychodynamic therapy and cognitive-behavioral therapy in social anxiety disorder: A naturalistic follow-up." American Journal of Psychiatry. 2014;171(10):1074–1082.

National Institute for Health and Care Excellence (NICE). Generalised anxiety disorder and panic disorder in adults: management. Clinical guideline CG113. NICE, 2011. Also: Post-traumatic stress disorder. Guideline NG116. NICE, 2018.

Shedler J. "The efficacy of psychodynamic psychotherapy." American Psychologist. 2010;65(2):98–109.