If you're in crisis, this page is the wrong place. Samaritans, Ireland and the UK: 116 123, free, 24 hours. At immediate risk: A&E, or 999 or 112.
This is ordered by the strength of the evidence, not by what we stock. Supplements come fifth. The order is not flattering to us and we've left it as it is — a list like this is only worth reading if the person writing it had nothing to gain from the ranking.
One thing to settle first: this is about anxiety as a disorder. If what you have is ordinary anxiety about a specific situation, most of the heavy interventions below are unnecessary. The distinction is here, and it's worth five minutes before reading on.
1. Psychological therapy
Cognitive behavioural therapy has the strongest evidence base of anything on this page for most anxiety disorders, and it has one property medication does not reliably match: the effects tend to persist after treatment ends. You are learning something rather than taking something.
What it actually involves
People expect talking about their childhood. CBT for anxiety is closer to a training programme. It is structured, time-limited and active — typically weekly sessions over a few months, with work between them.
The central mechanism is worth understanding, because it explains why it works and why it's uncomfortable. Anxiety is maintained by avoidance. Avoiding the feared thing brings immediate relief, that relief reinforces the avoidance, and the prediction that something terrible would have happened never gets tested. CBT deliberately reverses that — approaching the thing, in graded steps, so the prediction can be disconfirmed by experience rather than argued with.
That is why "just think positively" is not what this is. You are collecting evidence, not adjusting your attitude.
The variants, because they aren't interchangeable
- Exposure-based CBT — the core approach for phobias, panic disorder and social anxiety, and the most effective option for specific phobia by a wide margin.
- Exposure and response prevention (ERP) — the specific treatment for OCD, and meaningfully different from generic anxiety CBT. Getting this one right matters; OCD is widely misunderstood, including by people who have it.
- Acceptance and commitment therapy, and metacognitive therapy — both with growing evidence, both working on your relationship to the worry rather than its content.
- Digital and guided self-help CBT — not a poor substitute. For mild to moderate anxiety the evidence is genuinely decent, particularly when there's some human support attached, and it is usually available immediately.
Access is the real barrier
Waits are long in most systems, and this is the single biggest practical obstacle to the most effective treatment available. Two things worth knowing. Self-referral is possible in more places than people realise — in the UK you can refer yourself to NHS Talking Therapies without going through a GP. And guided digital programmes can start while you're on a waiting list rather than instead of being on one.
2. Medication
SSRIs and SNRIs are first-line drug treatment for most anxiety disorders and work well for a substantial proportion of people.
Three things worth knowing before starting, because each of them is a common reason people stop too early:
- They take weeks, not days. Four to six weeks for a fair assessment, sometimes longer.
- They can briefly increase anxiety at the start. This is the point at which many people conclude it's making things worse and stop. Starting low and increasing slowly reduces it, and it typically settles.
- Don't stop abruptly. Discontinuation symptoms are real and unpleasant, and coming off is something to plan with whoever prescribed it.
Other options exist where those don't suit — pregabalin and buspirone among them, both with their own trade-offs, pregabalin including a dependence risk that is now well recognised. That's a prescriber conversation, not a comparison to make from a web page.
Benzodiazepines deserve a specific note. They work, quickly, which is exactly the problem. Tolerance develops, dependence follows, and withdrawal produces severe anxiety — so the drug that relieved the symptom becomes a cause of it. Current guidance restricts them to short-term crisis use rather than ongoing treatment of an anxiety disorder. If you've been on one for months, that is a conversation to have, and not one to act on alone: stopping abruptly can be dangerous.
Therapy and medication together
Combination often outperforms either alone, and the two are not competing philosophies. Medication can lower the volume enough to make the work of therapy possible, which is a good reason not to treat the choice as ideological.
3. Exercise
The evidence here is stronger than its reputation, and it costs nothing. Both aerobic exercise and resistance training show benefit for anxiety symptoms, higher intensity generally more than lower, and the effect appears within weeks rather than months. Which type, and at what intensity, we've gone through separately.
The mechanism plausibly includes something useful beyond fitness: exercise produces raised heart rate, breathlessness and sweating — the same bodily sensations that anxiety produces — in a context where they are obviously harmless. For panic disorder in particular, repeatedly experiencing those sensations without catastrophe is close to what exposure therapy does deliberately.
The honest caveat: "just exercise" is unhelpful advice delivered to someone whose anxiety is stopping them leaving the house. It works better as an adjunct than as a first instruction, and the useful version of the advice is smaller than the version usually given — a walk counts.
4. The behavioural inputs
Sleep. The relationship runs both ways and each worsens the other: anxiety disrupts sleep, and sleep deprivation measurably increases anxiety the following day. Breaking into that loop at the sleep end is often easier than at the anxiety end. If insomnia is the entry point, CBT-I has better evidence than anything you can buy — and if it's stress specifically that's wrecking your nights, that mechanism is worth understanding.
Caffeine. Genuinely underrated, and not merely as an aggravator — high intake can produce the whole picture. Racing heart, jitteriness, a sense of dread, disturbed sleep. Its half-life averages around five hours but varies enormously between people, partly for genetic reasons, so an afternoon coffee is a different event for different people. If you're anxious and drinking four coffees a day, that's a free experiment worth running — reduce gradually, since abrupt withdrawal produces headache and irritability that will muddy the result.
Alcohol. Reduces anxiety for a few hours and raises it for the following day, and regular heavy use produces a persistently anxious baseline that improves on stopping. A reliably bad trade for anyone anxious, and the most common form of self-medication there is.
Breathing and mindfulness practices. Real, and smaller than the app-store enthusiasm suggests. Slow breathing has a genuine physiological basis and is useful in the moment; mindfulness-based programmes have reasonable evidence for anxiety symptoms. Neither is a substitute for items 1 and 2 if what you have is a disorder — but both are free, safe, and available at 3am, which is not nothing.
5. Supplements, last and narrowest
Here is where we're honest about our own category.
Some botanicals have real trial evidence for anxiety symptoms. Chamomile has been tested in a placebo-controlled trial in diagnosed generalised anxiety disorder and beat placebo, with the investigators themselves calling the effect modest. Passion flower has a small pilot trial against a benzodiazepine, which had no placebo arm. Magnesium has evidence that is thinner than the marketing suggests — we've set out exactly how thin rather than leaving it at an adjective.
Two things matter more than any of that. The trial doses are well above what is in most products, ours included — the chamomile research used 500 to 1,500 mg daily of a standardised extract, and our Sleep Formula contains 100 mg. We've set that comparison out in full, including the table.
And: no supplement in our range, or anyone's, is a treatment for an anxiety disorder. If what you have is the disorder, items 1 to 4 are where the effect is. A supplement at number five is not a small version of the same thing — it is a different order of magnitude. The longer answer to "can supplements help at all" is here, and it is a qualified yes with the qualifications doing most of the work.
A note on interactions
If you take an antidepressant, check before adding anything. St John's Wort and 5-HTP both interact with serotonergic medication, and both are in our Sleep Support formula. St John's Wort additionally induces liver enzymes and reduces the blood levels of a long list of drugs — including hormonal contraceptives, anticoagulants, anticonvulsants and some HIV medication.
That's on the label, and we keep a full list of how our own products interact with mental health medication. It is still the sort of thing worth raising with a pharmacist rather than working out alone.
And one thing to rule out before any of it
An overactive thyroid produces anxiety, tremor, palpitations and insomnia, and is regularly treated as an anxiety disorder for a long time first. It's one blood test. Details here, along with the short panel worth asking for if there's fatigue and poor concentration alongside.
This article is not medical advice. Treatment decisions for anxiety belong with a clinician who knows your history, and nothing here should be used to start, stop or change a prescribed medication.