Online vs In-Person Therapy for Anxiety: What Trials Compare

The question “can I do therapy online instead?” became urgent during the pandemic, but the evidence had been accumulating for years. Internet-delivered CBT for anxiety is one of the better-studied areas of digital health, and the findings are worth understanding in detail.

What’s being tested

Most of the evidence concerns structured internet-delivered CBT (iCBT): self-paced online modules covering the same content as face-to-face CBT (psychoeducation, cognitive restructuring, exposure, relapse prevention), typically delivered over 8–12 weeks.

This is different from video therapy, which is essentially the same as in-person therapy delivered over a screen. Most of the iCBT research tests a specific structured program, not a generic video session.

The critical distinction within iCBT is guidance:

Guided iCBT includes regular contact with a therapist — typically weekly emails, phone calls or secure messages — who reviews homework, provides feedback and encourages progression through modules.

Unguided iCBT provides the same content but without therapist contact. It is cheaper and more scalable, but adherence and outcomes are worse.

What the evidence shows

A Cochrane review of 38 trials found that therapist-supported iCBT produced large effects compared with waitlist controls and was not significantly different from face-to-face CBT in direct comparisons [2]. An updated meta-analysis of 20 direct comparison trials confirmed equivalent effects [3].

A systematic review specifically examining guided versus face-to-face delivery found no significant difference across 13 comparison trials [1]. An updated meta-analysis of computer-based therapy found clinically significant effects for both guided and unguided programs, but guided versions consistently performed better [4].

The guidance effect is substantial. A systematic review found that therapist guidance approximately doubled the effect size of internet interventions [6]. Unguided programs work, but less well and with higher dropout rates.

Where the evidence is strongest

Social anxiety disorder. iCBT for social anxiety has the most robust evidence base, with multiple large trials showing effects comparable to face-to-face CBT.

Generalized anxiety disorder. Strong evidence from multiple trials and Cochrane reviews.

Panic disorder. Good evidence, though fewer direct comparisons with face-to-face.

Specific phobias. Limited iCBT evidence, particularly for phobias that require real-world exposure (heights, animals) that may be difficult to arrange online.

Mixed anxiety. Several programs target transdiagnostic anxiety, with promising results.

Real-world implementation

Trial results don’t always translate to routine practice. Data from established iCBT services in several countries suggest that real-world outcomes are somewhat smaller than trial outcomes but still clinically meaningful [5]. Completion rates in routine services are often 50–60%, compared with 70–80% in trials — a gap, but smaller than you might expect.

The scalability advantage is significant. A therapist delivering iCBT can support 5–10 times as many patients as in face-to-face practice, because the modules do most of the psychoeducation and the therapist provides targeted feedback rather than full session delivery.

Limitations to consider

Not for everyone. Some people need or prefer the accountability, nuance and relational aspects of face-to-face therapy. Severe anxiety, active suicidality and complex presentations may require in-person care.

Exposure challenges. Disorders requiring in-vivo exposure with therapist support (severe agoraphobia, some specific phobias) are harder to treat online. The therapist can’t accompany the patient into feared situations.

The technology barrier. Older adults, people with low digital literacy, and those without reliable internet access may be excluded. This limits the equity argument for iCBT.

Therapeutic relationship. Online delivery changes the therapeutic relationship. Whether this matters for outcomes is debated — the outcome data suggests it doesn’t matter much on average, but averages mask individual variation.

Questions to ask about online therapy research

1. Is the study testing structured iCBT modules or video sessions?
2. Was the program guided or unguided?
3. Was the comparison face-to-face therapy, waitlist, or something else?
4. What was the completion rate, and how does it compare with face-to-face?
5. Were outcomes measured with the same tools across conditions?

What remains uncertain

Whether video therapy (synchronous, unstructured) is equivalent to in-person therapy needs more rigorous evidence — most of what exists is convenience data from the pandemic shift. Whether iCBT works equally well across cultures, languages and healthcare systems is being studied but not yet established. Whether the long-term effects of iCBT match face-to-face therapy is poorly studied. And whether fully automated AI-delivered therapy will eventually match guided human-supported programs is speculative.

Exploring online options. If in-person therapy isn’t accessible or preferred, guided internet-based CBT is an evidence-based alternative for most anxiety disorders. Look for programs with therapist support. If anxiety is affecting your daily life, talk to a GP or mental health professional. If you are in crisis or thinking about harming yourself, contact emergency services or a crisis line now: in the US, call or text 988; in the UK and Ireland, Samaritans on 116 123.