Route · Device-based

Microwave for sweating: response 89% versus 54% — the second number is the sham group

In a randomised, sham-controlled trial, the responder rate 30 days after treatment was 89% (72/81) active and 54% (21/39) sham (P < .001). That means more than half the people who received a fake procedure reported their sweating had dropped. Efficacy held at 74% by three months and 69% at twelve.

Last checked Sep 19, 2026

Korean classification Being checkedThe regulatory class has not yet been confirmed against a primary document. It will be filled in once it is.

This procedure delivers microwave energy to the underarm and destroys sweat glands with heat. In Korea it is best known by the device name miraDry.

There is already a microwave page on this site — Onda. Same kind of energy, different target. Onda aims at fat and laxity; this aims at sweat glands. The evidence situations are opposites too: for Onda we found no randomised trial at all, while this has a sham-controlled trial.

But the most important number on this page is not the 89%.

It is the 54% in the sham group. It is the largest sham response rate we have seen anywhere on this site — larger than on LED, PRP or deoxycholic acid.

It is worth remembering that figure whenever you read a number from an uncontrolled advertisement.

This is not a cosmetic you apply. It is a clinical procedure.

What is established

Moderate evidence

89% active, 54% sham — a gap of 35 points

A 2012 randomised, sham-controlled, blinded study. The authors describe it as the first investigational device study of this procedure.

The entry criteria were demanding.

Primary axillary hyperhidrosis
A Hyperhidrosis Disease Severity Scale (HDSS) score of 3 or 4 — people whose daily life is affected
Baseline sweat production greater than 50 mg per 5 minutes

Not relying on a subjective scale alone but weighing the actual sweat to set the entry bar is what makes this trial's conditions solid.

A responder was defined as someone reporting an HDSS score of 1 or 2 after treatment.

Thirty days after treatment:

Active: 89% (72/81)
Sham: 54% (21/39)
P < .001

The difference is statistically clear. But the two numbers have to be read side by side.

What was the sham? Participants lay under the same device, went through the same preparation, and believed they had been treated. And more than half of them changed their answer, 30 days later, to 'my sweating does not trouble my daily life'.

Why would that be so large? A few reasons are guessable.

The endpoint is a self-rated scale. HDSS asks, on four points, how much sweating interferes with daily life. There is wide room for expectation.
Hyperhidrosis swings heavily with circumstance. Tension makes it worse; reassurance after a procedure can make it better.
Entry required a score of 3–4. People who enrol at their worst come down simply by returning to their own average.

The procedure's share is not the 89% but the gap between 89% and 54%. The botulinum toxin page sets out the same calculation — there it was 75% against 25%.

Comparing the two is interesting. The raw response rate is higher for microwave (89% against 75%), but the gap over placebo or sham is larger for botulinum (50 points against 35). The endpoints and populations differ, though, so they cannot be compared directly.

Randomised controlled trial in people · 120 participants · 52 weeks · Compared with vehicle (same formula minus the active) · Funding manufacturer-funded · Endpoint clinical events (cancers, lesion counts) PMID 22289389

Early research

It held to twelve months — but there is no sham group in that stretch

How long an effect lasts is the question we have written 'not established' against on most procedure pages here.

This trial has an answer.

Three months: 74%
Twelve months: 69%
In the authors' words, efficacy was stable through twelve months, when follow-up ended

After falling from 89% at 30 days to 74% at three months, it barely moved from there to twelve. An early drop, then flat.

A large condition attaches.

The sham group was followed for six months; the active group for twelve.

So the 69% at twelve months has nothing to compare against. The deoxycholic acid page records the same problem — there is three-year follow-up data, but no control group across it.

Why it happens is understandable. Holding people on a sham procedure for a year is impractical and ethically difficult. Even so, the numbers have to be read knowing that.

And here the 54% from the item above matters again. We could not verify from the abstract what the sham group's responder rate looked like at six months. If it also stayed high, the meaning of 69% at twelve months changes.

So this item sits at 'early evidence' with the comparator marked 'none'. It should not be read with the same weight as the item above.

Randomised controlled trial in people · 81 participants · 52 weeks · Compared with no control group (before-and-after only) · Funding manufacturer-funded · Endpoint clinical events (cancers, lesion counts) PMID 22289389

Not established

This is the one randomised controlled trial

The only randomised controlled trial we verified for this procedure is the one above.

And the authors themselves wrote:

As with any new procedure, findings from this first investigational device study identified optimization strategies for the future.

What 'first' means on this page. First trials generally produce better results than the ones that follow. The protocol is not yet refined, but it is also the trial run by the most invested team on the most suitable participants.

By this site's rule, 'strong evidence' requires two or more distinct randomised trials. Here there is one, and it is manufacturer-funded. Deoxycholic acid had two placebo-controlled phase 3 trials and still did not rise, for the funding reason.

We have also registered a 2022 systematic review of palmar hyperhidrosis. It covers palms rather than underarms, so it does not apply directly to the procedure above, but it stays in this page's sources as a picture of the evidence landscape for hyperhidrosis treatment generally.

Separating what is and is not established:

It significantly beat sham at 30 days — established
The gap was 35 points — established
It held to twelve months — established, but without a control
Whether another research group produced the same result — not established
Whether it beats botulinum toxin head to head — not established

Known risks

In the trial above, adverse events were generally mild, and all but one resolved over time. What that one was is not in the abstract.
Swelling, pain, altered sensation and lumps in the underarm are reported.
This destroys tissue with heat. Nerves run through the underarm, and temporary arm tingling or weakness has been reported. Check directly with your clinician before the procedure.
Underarm hair is also reduced. That may or may not be what you want.
Sweat glands are held not to regenerate, so this is difficult to reverse. In that respect it differs in character from botulinum toxin, which wears off after some months.

What is not established

There is one randomised controlled trial, and it is manufacturer-funded.
There is no control group across the twelve-month stretch (second item above).
We could not verify the sham group's responder rate at six months from the abstract.
We found no randomised trial comparing it head to head with botulinum toxin. Both are used for axillary hyperhidrosis, and there is no answer on which to do first.
We found no evidence at this level for palmar or plantar hyperhidrosis.
How many sessions are appropriate — we found no established source.
We have not yet opened the original Korean device classification, so it is left as 'being verified'.

Whether to have it done is a decision to make with a doctor.

What this article relies on

PubMed — 미국 국립의학도서관 문헌 데이터베이스National Library of Medicine · Checked on 2026-09-16