Three options. One decision. Most people make it without knowing what the evidence actually shows about each: not what sounds reasonable, not what is easiest to access, but what actually works, for how long, and at what cost.
Cognitive Behavioral Therapy is the most extensively researched psychological treatment for anxiety disorders. Across hundreds of randomised controlled trials covering generalised anxiety disorder, social anxiety, panic disorder, health anxiety, and specific phobias, CBT consistently produces large effect sizes, with the majority of people experiencing significant symptom reduction.
What distinguishes CBT from the alternatives is not just that it works, but how it works. It targets the specific mechanisms that maintain anxiety: the threat appraisals, the avoidance behaviours, and the safety-seeking that keeps the loop running. Because it changes the underlying pattern rather than suppressing symptoms, the effects hold after treatment ends. Most people who complete CBT remain improved at one and two year follow-up without ongoing treatment.
Medication for anxiety, primarily SSRIs (selective serotonin reuptake inhibitors) and SNRIs, works by modulating neurotransmitter systems that regulate the threat response. It is effective, particularly in the short term, and for many people provides significant relief within 4 to 8 weeks. The evidence is real and substantial.
The critical limitation is what happens when medication is discontinued. Unlike CBT, medication suppresses symptoms rather than changing the underlying mechanisms. When the medication stops, the mechanisms remain. Relapse rates after discontinuation are substantially higher than after CBT, and many people find themselves on medication indefinitely to maintain the improvement.
Medication is most clearly appropriate as a complement to therapy when anxiety is severe enough to make therapeutic engagement difficult, or as a bridge while waiting for therapy to take effect. As a standalone long-term treatment, the evidence for therapy is stronger on the durability dimension that most people care about most.
Self-help approaches, books, apps, online resources, and self-directed practice of CBT techniques, produce real improvements in anxiety symptoms. Multiple meta-analyses confirm that self-help for anxiety is significantly better than no treatment. The effect sizes are genuine, just smaller than those produced by guided therapy.
The limitation of self-help is not that the techniques do not work. They are often the same techniques used in therapy. The limitation is that applying them consistently without accountability, without feedback on whether you are doing them correctly, and against the avoidance pull that anxiety produces, is significantly harder than doing so with a therapist's guidance. The dropout rates for self-directed programmes are substantially higher than for guided therapy.
Self-help works best for milder anxiety, as a supplement to therapy between sessions, or for maintaining gains after therapy ends. For moderate to severe anxiety, it is rarely sufficient as a standalone approach.
Note: This guide is for informational purposes only and does not constitute medical or clinical advice. Medication decisions should always be made with a qualified doctor or psychiatrist. Some links on this page are affiliate links.