Yes. For treatment-resistant depression, roughly 50 to 65 percent of patients respond to a full TMS course and about one third reach remission, in sham-controlled trials and large registries. That is substantially better than another medication trial at the same stage. Evidence is also solid for OCD and anxious depression, and thinner for other uses.
Patients ask me this question in almost every TMS consultation, usually in one of two forms. The hopeful form: "Does this actually work?" The skeptical form: "Is TMS a hoax?" Both deserve a straight answer built on data rather than brochures, so here is what the evidence says, where it is strong, and where it is not.
The short answer
For major depression that has not responded to medication, TMS works for a majority of patients. Across controlled trials and real-world registries, about half to two thirds of patients respond, meaning their depression scores fall by at least 50 percent, and roughly one third reach remission, meaning they no longer meet criteria for depression. Those figures come from people who had already failed an average of two to four antidepressants, which makes them more impressive than they look at first glance.
The controlled trials
TMS earned its first FDA clearance in 2008 on the strength of a multi-site randomized trial in which patients received either real TMS or a sham coil that looked and sounded identical. The active group had meaningfully higher response and remission rates at six weeks. A second, independent trial funded by the National Institute of Mental Health, often called the OPT-TMS study, repeated the design with a stricter sham and found the same thing: active TMS roughly tripled the odds of remission compared with sham.
Deep TMS, the BrainsWay H-coil system used at our Palm Beach Gardens office, was cleared in 2013 after its own multi-site sham-controlled trial. In that study, about one third of patients on active Deep TMS reached remission at five weeks, compared with roughly half as many on sham, in a group that had failed multiple medications. That trial also tracked patients through a maintenance phase, which is where some of the durability data comes from.
The sham design matters. A sham coil produces the same clicking and scalp sensation, so patients cannot easily tell which they received. The gap between active and sham is the part that cannot be explained by expectation or the daily attention of coming to a clinic.
The real-world registries
Controlled trials enroll carefully selected patients. Registries follow everyone treated in ordinary clinics, which is closer to what you would experience. The largest ones, including a multi-site NeuroStar outcomes registry with several thousand patients and BrainsWay's post-marketing data, report response rates in the range of 58 to 70 percent and remission rates of roughly 30 to 45 percent on clinician-rated scales.
Why are registry numbers often higher than trial numbers? Partly because clinicians in practice adjust protocols, extend courses and treat alongside medication, which trials restrict. Partly because of some optimism in open-label ratings. The honest range to plan around is the one I give patients: a bit better than one in two for response, about one in three for remission.
How TMS compares with the alternatives at this stage
The right comparison is not TMS versus a first antidepressant. It is TMS versus whatever else you would do after two or more medications have failed.
The STAR*D trial followed people through up to four sequential medication steps. Roughly a third remitted on the first antidepressant, about a quarter of the remainder on the second step, and by the third and fourth steps remission rates fell to the low teens. By that point TMS, with a remission rate near one in three, outperforms another switch or augmentation.
Against ECT, TMS is less effective for the most severe and psychotic depressions, where ECT remains the standard. For moderate to severe non-psychotic depression, several comparisons find the two closer than most people expect, and TMS wins clearly on side effects.
Against Spravato, there are no large head-to-head trials. Both work in treatment-resistant depression; Spravato tends to act faster, TMS tends to produce longer medication-free remissions. We offer both, and the choice is individual.
Who does best
Predictors of good TMS outcomes, consistent across studies, include:
- Fewer failed medication trials. People who have failed one or two medications respond better than those who have failed six.
- Shorter current episode. Depression that has lasted under two years responds better than chronic, decade-long episodes.
- Younger age, though older adults still respond and Medicare covers the treatment.
- Completing the full course. Early responders exist, but a meaningful share of responders do not turn the corner until weeks four to six.
- Continuing an antidepressant during treatment, in most studies.
Predictors of weaker outcomes include active substance use, untreated sleep apnea, and severe personality pathology. None of these are absolute exclusions, but they are reasons to address other factors in parallel.
Beyond depression
OCD
Deep TMS was cleared for OCD in 2018 after a multi-site sham-controlled trial in which about 38 percent of patients on active treatment responded, versus about 11 percent on sham, among people who had already failed medication and therapy. OCD protocols use a different coil target and involve brief symptom provocation before each session. The evidence here is good but the response rate is lower than in depression, and TMS works best alongside exposure and response prevention therapy.
Anxious depression
In 2021 the FDA cleared Deep TMS for anxious depression, based on pooled trial data showing that anxiety symptoms improve alongside depressive symptoms in patients with major depression. This is not the same as a clearance for generalized anxiety disorder on its own, and I tell patients that plainly. If your primary diagnosis is anxiety without depression, the evidence is promising but preliminary.
Smoking cessation
Deep TMS was cleared for short-term smoking cessation in 2020, based on a trial showing higher quit rates at four months than sham. It is used far less often than the depression protocol.
Other uses
TMS is studied for PTSD, bipolar depression, chronic pain and cognitive symptoms. Results are mixed, trials are small, and insurance does not cover these uses. I will tell you when a use is off-label and what the evidence actually supports.
Is TMS a hoax?
No. The skeptic's position usually rests on three points, so let me address them.
First, "it is just placebo." Sham-controlled trials with blinded raters consistently show active TMS beating sham, and the NIMH-funded OPT-TMS trial was designed specifically to rule this out.
Second, "the effect sizes are small." The difference between active and sham in trials is modest in absolute terms, which is true of almost every treatment in psychiatry. The clinically relevant number is the remission rate in people who had already failed several medications, and that number is good.
Third, "clinics oversell it." Some do. Any clinic quoting 80 or 90 percent success is quoting the best registry response figure, not remission, and not in a treatment-resistant population. If you hear that, ask for the remission rate and the definition.
What "working" looks like in practice
We measure depression every two weeks during the course with a standardized scale, usually the PHQ-9, and with your own account. A typical response begins as better sleep and energy around week three, then improved mood and interest around week four or five. Some patients plateau partway, and we adjust the protocol. A minority feel nothing by week six; for them we discuss extending the course, switching the protocol or moving to Spravato or another option.
Bottom line
TMS is one of the better-evidenced interventions in psychiatry for depression that has resisted medication. It is not a cure and it does not work for everyone, but the odds are substantially better than another pill at the same stage, and the side-effect profile is the mildest of any major option. Anyone who has failed two antidepressants should at least have the conversation.
Frequently asked questions
What is the TMS success rate?
For treatment-resistant depression, about 50 to 65 percent of patients respond (symptoms cut at least in half) and roughly one third reach remission. Registry data from ordinary clinics sometimes reports higher response figures. Rates for OCD are lower, around 38 percent response in the pivotal Deep TMS trial.
How soon will I know if TMS is working?
Most responders notice change in the third or fourth week, often as better sleep and energy before mood improves. A meaningful share do not respond until weeks five or six, so completing the course matters. We track symptoms with rating scales every two weeks.
Does TMS work for anxiety?
Deep TMS is FDA-cleared for anxious depression, meaning anxiety symptoms that accompany major depression. For anxiety disorders without depression, early studies are encouraging but the evidence is not yet strong enough for a clearance, and insurance does not cover that use.
Does TMS work better with or without medication?
Most trials and registries show slightly better outcomes when patients continue an antidepressant during TMS, and continuing medication afterward helps maintain the response. Your psychiatrist will decide with you whether to keep, adjust or taper medication around the course.
Related reading
- What Is TMS Therapy? A Plain-English Guide
- How Long Do TMS Results Last?
- TMS vs ECT: Comparing Two Brain Stimulation Treatments
- Treatment-Resistant Depression: What to Try Next
This article is educational and does not replace evaluation by a licensed clinician. If you are in crisis, call or text 988.