Can Anxiety Ever Go Away?
Not exactly, and that is better news than it sounds. Anxiety is a normal function, not a defect, so the realistic goal is not deleting it but getting it back to a level that is proportionate and useful. Anxiety disorders, on the other hand, do respond to treatment, and long-term follow-up studies find substantial numbers of people in remission years later. What the research also shows is that untreated anxiety tends to be persistent, and that residual symptoms are the strongest predictor of it coming back.
Two different questions hiding in one
When people ask whether anxiety ever goes away, they usually mean one of two things. Either "will I stop being an anxious person", or "will this state I am in right now ever end". Those have different answers.
The first is largely the wrong target. Anxiety is a threat-detection system, and everyone has one. A person with a sensitive alarm system is often also the person who notices things early, prepares properly and takes other people seriously. Telling that trait apart from a disorder is worth doing early, and stress versus anxiety covers the nearest confusion. Aiming to remove that entirely is aiming at something that would not actually be good for you, and chasing it tends to make people more anxious, because now there is a deadline.
The second question, about the disorder rather than the trait, has a genuinely encouraging answer.
What happens if anxiety is left alone
Untreated anxiety disorders are more persistent than most people expect. In the Harvard-Brown Anxiety Research Program, 60 per cent of participants with generalised anxiety disorder still had active GAD after two years, and at five years 66 per cent of that group had only partial remission or none. Epidemiological Catchment Area data found 40 per cent had carried the diagnosis for more than five years, and of those who did remit, 27 per cent relapsed within three years.
General population research on recurrence found 23.5 per cent within two years of baseline, with no meaningful difference between panic disorder, social anxiety disorder and GAD.
The practical reading is that waiting it out is not a strong plan. Anxiety disorders are not usually self-limiting in the way an acute stress reaction is, largely because avoidance quietly teaches the system that the avoided thing was genuinely dangerous. Every time you swerve the presentation, the phone call or the social event, the alarm gets confirmation. That mechanism is precisely what graded exposure work is built to reverse.
What happens with treatment
Long-term follow-up data on CBT for anxiety is unusually good, because researchers have gone back to these samples years later.
A 2025 follow-up of 93 people treated for GAD, two to eight years after treatment, found gains maintained on almost all measures, with 57 to 77 per cent categorised as recovered. A ten-year follow-up of a randomised controlled trial in older adults found 58 per cent remission of all diagnoses after CBT against 27 per cent in the control condition, 67 per cent remission of the primary diagnosis against 42 per cent, and relapse of 25 to 31 per cent against 50 to 78 per cent. That study concluded around 68 per cent of people treated with CBT were in remission of their anxiety disorder up to ten years later.
One finding in that study is worth more than all the headline percentages. People who showed a clear response by the end of treatment were seven to nine times more likely to be in remission a decade later than those who finished with residual symptoms still present.
Longer-term data is not uniformly rosy. An eight to fourteen year follow-up found 50 per cent markedly improved, of whom 30 to 40 per cent were recovered, with poor outcome for 30 to 40 per cent. Recovery rates also swing widely depending on the threshold used; on strict worry-based criteria the figures are more modest, closer to 46 per cent at end of treatment and 57 per cent at twelve months.
The finding that should change what you do
Take those two things together. Residual symptoms at the end of treatment predict relapse, and most relapse happens in the first two years. That has a very practical implication, and it is the one I would most want you to take from this page.
Stopping therapy at the point where you feel noticeably better is the single most common way people end up back here in three years. That point tends to arrive early, often once sleep improves and the worst of the panic settles, and it is enormously tempting. But it is usually the point at which the avoidance has only partly been dismantled and the underlying rules are still intact.
This is where the work I do differs from what people expect. Most of what makes anxiety treatment stick is not in the first phase, where symptoms drop. It is in the unglamorous later phase, where we go after the things you are still quietly avoiding, the reassurance-seeking you have not mentioned, and the beliefs that are still running underneath. I am a registered counsellor and psychotherapist, ACA Level 2 and PACFA, working online worldwide and in person in Melbourne, and a good deal of what I do with anxious clients is resisting the urge to finish early.
Depending on the picture that might be CBT for the thinking patterns and the avoidance, graded exposure work where specific situations have been ruled out of your life, or mindfulness-based cognitive therapy, which was originally developed to prevent relapse and is well suited to people who improve and then slide back. You can read about how I approach anxiety or arrange a first session.
So what does recovered actually look like
Not a life without nerves. What people describe when it has gone well is that anxiety shows up in proportion to the situation and then leaves, rather than arriving unprompted and staying for days. Sleep is normal most nights. The list of avoided things stops growing and starts shrinking. A bad week is a bad week rather than the beginning of the end.
Anxiety returning under real pressure, a bereavement, a job loss, a health scare, is not failure. It is the system doing its job. The difference after good treatment is that it comes down again afterwards. The same prognosis question for trauma is answered in does PTSD ever go away.
References
- The long-term clinical course of generalized anxiety disorder. Journal of Clinical Psychiatry. View source
- Long-term remission and relapse of anxiety and depression in older adults after cognitive behavioural therapy: a 10-year follow-up of a randomised controlled trial. Journal of Affective Disorders, 2024. View source
Written by Shagoon Maurya
Registered counsellor and psychotherapist, ACA Level 2 and PACFA. Post graduate studies in Counselling & Psychotherapy, University of Adelaide, Australia. Working online worldwide and in person in Melbourne.
Anxiety, recovery and relapse
Anxiety as a function does not go away, and would not be good for you if it did, since it is the threat-detection system everyone has. An anxiety disorder is different and does respond to treatment. Long-term follow-up of CBT found around 68 per cent of people in remission of their anxiety disorder up to ten years later, against 27 per cent in a control condition.
Often not. In the Harvard-Brown study, 60 per cent of people with generalised anxiety disorder still had active GAD at two years, and 66 per cent of that group had only partial remission or none at five years. Avoidance is a large part of why: each time a feared situation is swerved, the system gets confirmation that it was genuinely dangerous.
Follow-up studies are encouraging. A 2025 study of people treated for GAD two to eight years earlier found 57 to 77 per cent categorised as recovered. A ten-year randomised trial follow-up in older adults found 58 per cent remission of all diagnoses after CBT against 27 per cent in the control group, with relapse of 25 to 31 per cent against 50 to 78 per cent.
The strongest predictor is residual symptoms at the end of treatment. In ten-year follow-up data, people who showed a clear response by the end of therapy were seven to nine times more likely to be in remission a decade later than those who finished with symptoms still present. Most relapse occurs within the first two years.
Feeling noticeably better is usually not the right stopping point. It tends to arrive once sleep improves and acute symptoms settle, which is often before the avoidance has been fully dismantled or the underlying beliefs addressed. Finishing at that point is one of the most common routes back into treatment a few years later.
Anxiety that shows up in proportion to the situation and then leaves, rather than arriving unprompted and staying for days. Normal sleep most nights. A list of avoided situations that shrinks rather than grows. Anxiety returning under genuine pressure such as bereavement or job loss is not relapse; the difference after good treatment is that it settles again afterwards.