If you walk into any therapist’s waiting room in Chicago, Sydney, or London, there’s a good chance the person there has received CBT training. It will most likely be identified by the diplomas on the wall. It may be described in the intake paperwork. It has become so ingrained in Western mental health care that challenging it can feel almost rude, similar to voicing concerns about antibiotics during a medical conference. However, a thorough reading of the research reveals a more nuanced narrative than the one typically summed up in clinical brochures.
Certain aspects of CBT have been proven beyond a reasonable doubt by two decades of thorough meta-analysis. Large, consistent effect sizes for major depressive disorder, generalized anxiety, OCD, and PTSD were found in a 2021 umbrella review published in World Psychiatry that looked at 269 different meta-analyses. That discovery is not insignificant. A therapy deserves its first-line status if it is effective for a wide range of conditions, in populations ranging from adolescents to veterans, across cultures and delivery formats, including digital. The neuroimaging data is especially striking: successful CBT courses result in quantifiable changes in the brain’s threat-processing center, the amygdala, indicating that the therapy is altering not only how people discuss their problems but also how their brains react to them. Compared to self-reported mood scores, that is more difficult to ignore.
But when you go from “does it work” to “does it work for everyone, durably, and better than alternatives,” the whole picture of what CBT proves becomes less clear. Approximately half of adults treated with a standard brief CBT course either relapse within months or fail to achieve full remission, according to research published in major psychiatric journals. According to some estimates, that percentage is even higher for depression in particular—roughly 75% of patients do not sustain the benefits of short-term treatment. This is not a small footnote. Nearly half of the target audience is represented. Reading the literature closely gives the impression that the field loudly boasted about CBT’s achievements while viewing its non-responders as a technical issue that would be resolved in the future.
Additionally, the comparison data is more humble than most people realize. CBT does not consistently outperform other active psychotherapies, such as psychodynamic therapy, humanistic approaches, or even non-directive counseling, in numerous controlled trials. In psychotherapy research, the “Dodo bird verdict”—named after Alice in Wonderland’s statement that “all have won and all shall have prizes”—has steadfastly resisted attempts by researchers to refute it. The benefits of CBT may lie less in its particular methods and more in the fact that it is measurable, goal-oriented, and structured—qualities that make it easier to research and train therapists in, which in turn produces more evidence, which leads to more endorsements. The therapy’s predominance in guidelines may reflect both its superiority and its researchability.
Complex, long-standing presentations are where CBT exhibits its most obvious limitations. A twelve-session thought record program is frequently insufficient for severe personality disorders, deep relational trauma accumulated over years, or situations where the body manifests the distress as clearly as the mind. In trauma-focused settings, practitioners are increasingly using somatic approaches, EMDR, DBT, or Schema Therapy to address issues that CBT by itself frequently fails to address. That is more of a misuse than a failure of CBT, but the misuse has been pervasive enough to be significant. Many people went through years of CBT referrals without anyone pointing out that they might benefit more from a different strategy.
Despite decades of protocol improvement and significant research investment, it is difficult to ignore the fact that effect sizes in anxiety treatment have essentially plateaued. The ceiling might just be a reflection of cognitive restructuring’s limitations as a mechanism—the notion that recognizing and correcting false beliefs has a limited scope, beyond which other biological, social, or relational issues must be addressed. This contributes to the rapid growth of third-wave CBT approaches over the past ten years, including Dialectical Behavior Therapy, Acceptance and Commitment Therapy, and Mindfulness-Based Cognitive Therapy. These are efforts to expand the model into areas where its initial presumptions begin to falter.
CBT is not a failed project because of any of this. The structural approach is effective for a significant percentage of individuals, the evidence base is solid, and the long-lasting coping skills it imparts are useful long after the last session. However, the research’s actual findings are more limited and conditional than what decades of clinical enthusiasm have occasionally implied. CBT is effective. Just not for everyone, not always long-lasting, and not always superior to its predecessors. At this point, understanding that distinction seems to be a fundamental professional obligation.