The short answer: Exercise reduces anxiety symptoms, but it does not work as well as medication or CBT for diagnosed anxiety disorders. Meta-analyses of randomized trials put the effect of exercise at about 0.4 standard deviations, a small to moderate benefit. In head-to-head trials, clomipramine and group CBT beat exercise on most measures. Exercise is a strong add-on to treatment and a reasonable first step for mild symptoms, not a substitute for care when anxiety is moderate or severe.

What do randomized trials show about exercise for anxiety?

Several meta-analyses have pooled the trials. The numbers agree.

By convention, 0.2 is a small effect, 0.5 is medium, and 0.8 is large. Exercise lands between small and medium.

How does exercise compare with SSRIs and CBT?

Few trials have tested exercise directly against a drug or therapy. The ones that exist favor the established treatment.

  • The first direct comparison randomized 46 outpatients with moderate to severe panic disorder to 10 weeks of running, clomipramine 112.5 mg per day, or placebo. Both beat placebo. The drug improved symptoms earlier and more effectively. Dropout was 31% with exercise and 0% with clomipramine.
  • A 2010 trial of 75 outpatients with panic disorder paired paroxetine or placebo with aerobic exercise or relaxation training. Paroxetine beat placebo. Exercise did not differ from relaxation on most measures.
  • A 2013 trial of 36 people with panic disorder compared 12 weeks of group exercise (three sessions a week) with weekly group CBT. At 12 months, 52.6% of the CBT group showed clinically significant change, versus 11.8% of the exercise group.

A 2015 meta-analysis of 234 randomized trials (37,333 patients) reported pre-post effect sizes by treatment type. Medications averaged a Cohen's d of 2.02. SSRIs scored 2.09 and SNRIs scored 2.25. Individual CBT scored 1.30. Exercise scored 1.23, based on only three study arms. Placebo pills scored 1.29, so pre-post numbers include a large placebo effect. A 2014 systematic review of eight trials reached the same conclusion: exercise reduces anxiety but is less effective than antidepressants.

One caution. The 2023 umbrella review noted that its exercise effects were comparable to placebo-controlled effects reported for psychotherapy and drugs (SMD -0.22 to -0.37). Those figures come from different populations and comparison groups. When the same patients are randomized to exercise or a proven treatment, the proven treatment wins.

Which type and dose of exercise helps most?

A 2026 network meta-analysis in BMJ Open Sport & Exercise Medicine pooled 30 trials with 1,421 participants. In people with diagnosed anxiety disorders, resistance training had the largest effect (SMD -0.79, 95% CI -1.18 to -0.40). Mind-body exercise such as yoga scored -0.71 and aerobic exercise scored -0.65. The average program used 43-minute sessions, three times a week, for eight weeks.

Intensity matters. In the 2018 meta-analysis, high-intensity programs beat low-intensity programs with an SMD of -0.38 (95% CI -0.68 to -0.08). The 2023 umbrella review also found larger improvements with higher intensity. The 40-trial review found bigger effects when sessions lasted more than 30 minutes (0.36) and when programs ran 3 to 12 weeks (0.39).

These doses match the HHS Physical Activity Guidelines. Adults need 150 to 300 minutes of moderate aerobic activity each week, plus muscle-strengthening activity on at least 2 days.

How fast does exercise reduce anxiety?

Some relief is immediate. The HHS guidelines state that a single bout of physical activity can reduce anxiety. That effect is short-lived. Lasting change takes weeks. The trials above ran 8 to 12 weeks before measuring outcomes. In the panic disorder trial, clomipramine improved symptoms earlier than running did. That matches what the National Institute of Mental Health says about drugs: SSRIs and SNRIs may take several weeks to start working, and buspirone needs 3 to 4 weeks. Our guide on how long anxiety medication takes to work covers the drug side in detail.

What are the limits of the evidence?

The trials are small. A 2015 systematic review in Annals of Behavioral Medicine found only 12 randomized trials in anxious adults. Half had fewer than 50 participants and only one had more than 100. Dropout averaged 18.4% and reached 52% in one trial. The authors concluded that the evidence was "not of sufficient scientific rigor" to recommend exercise as a treatment for clinically elevated anxiety.

Newer reviews have the same problems. In the 2026 network meta-analysis, only 4 of 30 trials had a low risk of bias. Only 12 reported any follow-up, and none past 12 weeks. Most trials compared exercise with a waiting list, which inflates effects. In the 2015 treatment meta-analysis, waiting lists produced a pre-post effect of only 0.20, versus 1.29 for a placebo pill.

Evidence in young people is thinner. A Cochrane review of 16 trials with 1,191 participants aged 11 to 19 found a trend toward less anxiety with exercise (SMD -0.48, 95% CI -0.97 to 0.01). The result was not statistically significant.

When is medication still needed?

The American Psychiatric Association lists CBT as a first-line treatment and SSRIs and SNRIs as first-line medications. For moderate to severe cases, it says medication is appropriate. NIMH adds that benzodiazepines can be used to manage severe GAD. Regular physical activity appears in the APA guidance as a healthy habit that supports recovery, not as a primary treatment.

Medication or therapy is needed when:

  • Anxiety blocks work, school, sleep, or relationships.
  • You have panic attacks or avoid places because of fear.
  • Symptoms persist after 8 to 12 weeks of regular exercise.
  • Depression is also present. Both often need treatment together; see depression.md.
  • You cannot keep up an exercise routine. Dropout hit 31% in the exercise arm of the panic trial.

Not sure where your symptoms fall? Read our guide on when to see a doctor for anxiety. If you have thoughts of harming yourself, call or text 988.

How do you combine exercise with medication or therapy?

This is where exercise performs best. A pilot randomized trial added home walking to group CBT for panic disorder, GAD, or social phobia. Compared with CBT plus education, the walking group had lower anxiety scores on the DASS-21 (regression coefficient -3.41, p<0.05). The 2014 systematic review found that exercise plus an antidepressant improved outcomes in panic disorder, and that exercise added to group CBT helped social phobia. NIMH states that exercise can "reduce anxiety symptoms and enhance the effects of psychotherapy."

A plan based on the trial doses:

  1. Pick resistance training, brisk aerobic exercise, or yoga. All three have evidence.
  2. Aim for three sessions a week of 30 to 45 minutes.
  3. Give it 8 weeks before judging the result. Track symptoms weekly.
  4. Keep taking prescribed medication. Do not stop or cut a dose without asking your prescriber.
  5. Our guide to anxiety management techniques that work covers breathing, exposure, and sleep.

The bottom line

Exercise reduces anxiety with a small to moderate effect, about 0.4 standard deviations. Resistance training, aerobic exercise, and yoga all help, and higher intensity helps more. In direct comparisons, medication and CBT beat exercise, and the exercise trials remain small. For mild anxiety, an 8-week program of three weekly sessions is a sound first step. For moderate or severe anxiety, exercise belongs alongside treatment, not in place of it.

Last updated: September 2026. This article is for informational purposes only and does not constitute medical advice. Talk with a licensed clinician before starting, changing, or stopping any anxiety treatment, and check with your doctor before beginning a vigorous exercise program.