PEMF Therapy: What It Is, and What the Trials Actually Show

Affiliate disclosure: This post contains affiliate links. See our affiliate disclosure for details.

PEMF therapy is a pulsed magnetic field applied to the body, and the reason anyone thinks it might do something is a piece of nineteenth-century physics: a magnetic field that changes induces an electric current in whatever conducts nearby, including you. Whether the currents a $1,000 mat induces are large enough to matter is the entire question, and the honest answer in 2026 is that it depends heavily on which condition you are asking about. Chronic low back pain has the best supporting evidence. Knee osteoarthritis, the most-studied use, came back mixed and short of clinical significance in the newest meta-analysis. Fibromyalgia has a well-run trial where the sham worked as well as the device.

We should be straight about what this page is and is not. We buy recovery hardware for this site and write about what we own, which is a barrel sauna, an ice bath and a red light panel. A PEMF mat is not among it, so there is no hands-on verdict here and we are not going to invent one. What there is: every study below was read from its own abstract or full text rather than from a summary of it, every device spec was pulled from the manufacturer's own page or product sheet on 25 August 2026, and we did the one piece of arithmetic that almost no page in this category does, which is checking whether the mats being sold on the strength of these trials run anywhere near the field strengths the trials used. They mostly do not.

If you already know the science and just want the hardware, the PEMF mat rankings have the full comparison. If you are working out where PEMF sits against everything else you could spend the money on, the home recovery setup guide is the wider view, and the grounding mat guide covers a cheaper category with a very similar evidence problem. More on us on the about page, and the rest of the category sits under recovery.

Last reviewed: August 2026

What PEMF actually is

Inside a PEMF device is a coil of copper wire. Current runs through it in pulses, switching on and off at some rate between roughly 1 and 100 times a second on a consumer unit. Each switch creates a magnetic field that rises and collapses, and by Faraday's law of induction, a collapsing magnetic field drives a small electric current in any nearby conductor. Tissue is a conductor. That induced current is the whole proposed mechanism, and everything else in the category, the gemstones, the infrared, the negative ions, is a different therapy bolted onto the same mat.

Two numbers describe any device honestly.

Frequency, in hertz, is how fast the field pulses. Every brand publishes this. The OMI Full Body Mat spans 1 to 99 Hz; HealthyLine's TAJ series runs 1 to 30 Hz with a default of 7.83 Hz.

Intensity, the peak field strength, is how hard it pulses. This is the number that separates a research device from a wellness product, and it is published far less consistently. Worse, it appears in three units that differ by factors of a hundred and a thousand. One gauss is 100 microtesla, or 0.1 millitesla. So a mat quoting "2 gauss" and a study quoting "10 mT" are 50 times apart, and nothing on either page tells you that.

Hold onto the conversion. It is the single most useful thing on this page.

The regulatory picture, which is three separate things

The word "FDA" gets used in this category to mean at least three different things, only one of which is approval.

Approved, in the full sense. Non-invasive bone growth stimulators are real regulated medical devices that use pulsed electromagnetic fields to promote healing in fractures that have failed to unite. They are prescribed and used under orthopaedic supervision. On 16 April 2026 the FDA published a final order reclassifying them from Class III to Class II subject to premarket notification, effective 18 May 2026, which lets manufacturers use the 510(k) pathway rather than full premarket approval. This is the legitimate clinical history that the consumer category borrows its credibility from.

Cleared, for a narrow indication. Some pulsed electromagnetic devices hold 510(k) clearance as nonthermal shortwave therapy. SofPulse, cleared in 2008 under product code ILX, is cleared for adjunctive use in the palliative treatment of post-operative pain and edema in superficial soft tissue. That is a specific claim about swelling after surgery, not a general license to treat pain.

Registered, which means nothing about the device. Establishment registration is an annual administrative filing by anyone who makes or distributes devices for the US market. The FDA is unusually blunt about it: an entry in the registration and listing database "does not denote approval, clearance, or authorization of that facility or its medical devices," and the agency has publicly told firms to stop issuing official-looking "FDA registration certificates" because using them to imply review misbrands the device.

Credit where it is due, HealthyLine states this correctly on its own FAQ page. Asked whether its mats are FDA-approved, the answer reads: "HealthyLine mats are FDA-registered as medical devices but not FDA-approved." A separate answer adds that the heating pads are "classified as a Medical II device and are 510K exempt." Oxford Medical goes further in the OMI product information sheet, which states that "the product is not a medical device; it is not intended to diagnose, treat, cure or prevent any disease."

When the manufacturer of the best mat in the category tells you in writing that it is not a medical device, that is the frame to buy in.

What the trials found, condition by condition

Three bodies of evidence matter for the reasons people actually buy these things.

Chronic low back pain: the strongest case

Sun and colleagues published a systematic review and meta-analysis in Clinical Rehabilitation in 2022 covering 14 randomised trials and 618 participants. PEMF beat placebo for pain relief, with a standardised mean difference of −1.01 (95% CI −1.42 to −0.6, p < 0.001). Splitting by duration, the effect held for chronic low back pain (SMD −0.6, 95% CI −0.94 to −0.25, p < 0.001) and did not reach significance for acute low back pain (SMD −0.46, 95% CI −0.99 to 0.07, p = 0.09). Physical function did not improve. Their conclusion is a fair summary of the whole field: beneficial for alleviating pain in chronic low back pain, no advantage for function.

Knee osteoarthritis: the most studied, and the least convincing

This is where most PEMF marketing points, and where the newest evidence is least flattering. Chang, Lin and Huang published a systematic review and meta-analysis in Medicina on 2 April 2026, searching PubMed, Embase and the Cochrane Library for trials from 2015 to 2025 and pooling nine randomised trials with 457 patients.

No significant improvement in VAS pain or total WOMAC score at one month. Underneath that headline the timing is genuinely odd: WOMAC pain improved significantly at 18 to 21 days (MD −1.63, 95% CI −2.43 to −0.82) but not at one month, while WOMAC stiffness (MD −1.11) and daily activity (MD −3.39) improved significantly only at one month. Objective functional measures did not improve. Overall risk of bias across the included studies was high.

Then the sentence that ought to appear on every page that cites this paper and appears on almost none of them. The authors list the minimal clinically important difference for each outcome, 7.8 points for WOMAC total, 8.7 for pain, 20.2 for stiffness, 14.5 for function, and 19.9 mm for VAS, and write that "our pooled MDs, even when statistically significant, did not meet these MCID thresholds." Statistically detectable, probably not perceptible.

That is consistent with the older meta-analysis it supersedes. Chen and colleagues, writing in the Journal of Rehabilitation Medicine in 2019 and pooling eight trials and 421 patients, found a benefit for physical function (WMD −5.28, 95% CI −9.45 to −1.11, p = 0.01) and nothing significant for WOMAC total, WOMAC pain, VAS pain or stiffness.

Fibromyalgia: the one everyone skips

Multanen and colleagues ran 108 women with fibromyalgia through a crossover trial published in Bioelectromagnetics in 2018: 12 weeks with an active BEMER device and 12 weeks with an identical inactive one, in random order. Pain fell significantly during the active period, by 12 points on a 100 mm scale. It also fell significantly during the sham period, by 11. Per-protocol analysis found no difference between active and sham on any outcome.

That is not a failed study. It is a well-designed study returning a clear negative, and it is a useful reminder of how much of the reported benefit in an unblinded home setting can come from lying still for 20 minutes a day.

One more finding worth carrying. A 2022 network meta-analysis of adjuvant treatments for fracture nonunion in BMC Musculoskeletal Disorders found electromagnetic field therapy carried a significantly higher adverse-effect risk than control (OR 13.21), on a very wide confidence interval of 1.58 to 110.40. That interval is too wide to draw much from, but "electromagnetic therefore harmless" is not something the literature actually establishes.

The number nobody checks: dose

Here is the arithmetic that changes how you read every study above.

Chang's 2026 review publishes the field strength and frequency each of its nine trials used. Set them against what a home mat produces.

Trial in the 2026 review	Field strength	Frequency	Versus a 2 gauss home mat
Wuschech, 2015	105 mT	4–12 Hz	525× stronger
Elboim-Gabyzon, 2023	10 mT	30 Hz	50× stronger
Pasin, 2025	10 mT	30 Hz	50× stronger
Comino-Suárez, 2025	10 mT	50 Hz	50× stronger
Yabroudi, 2024	5 mT	50 Hz	25× stronger
Ehsan Hashemi, 2025	1.2 mT	10–100 Hz	6× stronger
Wang, 2024	1 mT	50 Hz	5× stronger
Dündar, 2016	0.1 mT	50 Hz	half as strong
Bagnato, 2016	0.0001 mT	1000 Hz	2,000× weaker

A HealthyLine TAJ mat publishes a PEMF peak intensity of 2 gauss, which is 0.2 mT. The OMI Full Body Mat's product sheet gives 210 microtesla at 8 Hz and 220 microtesla at 99 Hz, so 0.21 to 0.22 mT. Seven of the nine trials ran at 1 mT or above, between five and five hundred times the field a home mat produces. Only two ran at or below it, and one of those, Bagnato, was four orders of magnitude below.

Bagnato also deserves an asterisk, because it is not really the same technology. That trial used an ActiPatch, which its own paper describes as a pulsed radiofrequency energy device carrying a 27.12 MHz signal, pulsed at 1000 Hz in 100 microsecond bursts. The 1000 Hz in the table is the pulse rate, not the field. So the one trial sitting in the corner of this review where pain relief reached significance is a wearable radiofrequency patch, not a low-frequency coil mat. Whatever it found, it is not a result about the thing you would be buying.

The frequency side is just as awkward. Chang's subgroup analysis split the trials at 1 mT for amplitude and 100 Hz for frequency, and found the two effects landed in opposite corners: WOMAC pain relief was significant only in the low-amplitude, high-frequency subgroup, while stiffness and daily activity improved significantly only in the high-amplitude, low-frequency subgroup.

Now place the mats. At 0.2 mT they are low-amplitude, which fits the pain-relief corner. But the OMI tops out at 99 Hz and the HealthyLine at 30 Hz, so neither reaches the high-frequency side of that split. On the review's own parameters, consumer mats sit in the low-amplitude, low-frequency quadrant, which is the one combination the subgroup analysis found nothing significant for.

Three caveats, and they matter. No trial in that review tested a consumer mat, so this is a mismatch between parameters and not a test of these products. The authors call their own subgroup result "hypothesis-generating." And the low-amplitude, high-frequency arm rests on a single 60-patient trial, the radiofrequency patch described above. But it is still a more useful thing to know before spending $1,000 than a brand's claim that PEMF is "clinically proven," and nothing about the mismatch runs in the buyer's favour.

The 7.83 Hz question

Most PEMF mats default to 7.83 Hz, and the marketing reason given is the Schumann resonance, the electromagnetic resonance of the cavity between the Earth's surface and the ionosphere, which sits near that frequency. HealthyLine's spec sheet even notes the default is 7.83 Hz and is displayed on the controller as 8 Hz.

The physics is real. The therapeutic inference is not evidence. A frequency occurring in nature does not make it a treatment dose, and there is no body of controlled human trials establishing 7.83 Hz as therapeutically special. Note also that in the studies with the most encouraging results, treatment frequencies were 30 to 50 Hz and above, nowhere near it.

Treat a 7.83 Hz default as a sensible arbitrary starting point, not as a reason to choose one device over another.

What the hardware costs

If you want to run the experiment on yourself, this is what the entry point looks like. Prices are what we saw on 25 August 2026, on Amazon US and on each brand's own store.

Device	Price	PEMF frequency	Published intensity	Heat?
OMI Full Body Mat	~$1,050	1–99 Hz	210 µT at 8 Hz, 220 µT at 99 Hz	No
HealthyLine TAJ 7224	~$1,049	1–30 Hz, default 7.83 Hz	2 gauss peak	Yes, to 158°F

Prices last checked 25 August 2026 on Amazon US and on healthyline.com, and they move, so confirm the current price on the brand's own site before you buy. Neither of these, nor the shorter TAJ 6024 below, carried a promotional discount on the day we checked.

Two things to take from that table rather than the prices themselves.

The OMI Full Body Mat is the one that makes a personal experiment readable. It is pure PEMF, with no heating layer and no gemstones, so none of the price is buying you warm rock and nothing else in the session is competing to explain how you feel afterwards. Oxford Medical also publishes its intensity at two named frequencies rather than one round number, which is a small sign of a company willing to be checked. Its own product sheet setting sessions at 20 to 30 minutes and stating flatly that this is not a medical device is the right expectation to buy with.

The HealthyLine TAJ 7224 is the opposite proposition and it is worth being clear-eyed about why people like it. Six coils at a 2 gauss peak, sine waveform, a 20-minutes-on, 100-minutes-off cycle, and a 27 lb mat filled with crushed amethyst and rounded jade and tourmaline that heats to 158°F. Most of the price is the heat and the stones, not the PEMF hardware, and most of what a user will feel is the heat. It carries a 90-day money-back guarantee, which for a device you are testing on yourself is the most useful spec on the page. The TAJ 6024 is the same series 12 inches shorter, at ~$849 on both Amazon and healthyline.com, which is $200 less.

For the full ranking, including the two best-known mats we do not recommend and why, see the PEMF mat comparison.

Who should not use PEMF

This section is short because the list is short and firm.

Do not use PEMF with an implanted electronic device. HealthyLine's own FAQ states that "it is not recommended to use PEMF therapy if you have a pacemaker or any kind of defibrillator," and advises consulting your healthcare provider before using any of the therapies. The same reasoning covers implanted defibrillators, insulin pumps, neurostimulators and cochlear implants, since a pulsed magnetic field can interfere with device electronics.

Manufacturers also routinely advise against use during pregnancy. If you have epilepsy, an active cancer diagnosis or a transplanted organ, ask the clinician managing that condition, not the company selling the mat.

And on combination mats, remember which therapy actually carries risk. A 2 gauss magnetic field is very weak. A surface running at 158°F is not, and heat is the more plausible route to a problem in a home setting, particularly for anyone with reduced skin sensation.

Frequently asked questions

What is PEMF therapy?

PEMF stands for pulsed electromagnetic field therapy. A coil inside a mat, pad or ring switches on and off many times a second, producing a magnetic field that rises and falls rather than sitting still like a fridge magnet. A changing magnetic field induces a small electric current in anything conductive nearby, including tissue, and that induced current is the mechanism every PEMF claim rests on. Two numbers describe a device: frequency in hertz, and intensity in gauss, millitesla or microtesla. Home mats typically run 1 to 99 Hz at around 2 gauss, which is 0.2 millitesla.

Does PEMF therapy actually work?

It depends entirely on the condition. Chronic low back pain has the best case: a 2022 review of 14 trials and 618 participants found PEMF beat placebo on pain, though not on physical function, and the effect held only in chronic rather than acute cases. Knee osteoarthritis is weaker, with the 2026 meta-analysis of nine trials and 457 patients finding no significant improvement in VAS pain or total WOMAC at one month, and its authors noting that even the significant results missed the thresholds for clinical meaningfulness. Fibromyalgia has a crossover trial in which the sham device matched the real one.

Is PEMF therapy FDA approved?

Some PEMF devices are, and the mat you can buy online is almost certainly not one of them. Non-invasive bone growth stimulators are regulated medical devices, reclassified by the FDA from Class III to Class II in a final order published 16 April 2026 and effective 18 May 2026. Some pulsed shortwave devices hold 510(k) clearance for post-operative pain and swelling. Consumer mats are neither. "FDA registered", which brands sometimes lean on, is an administrative filing that the FDA says does not denote approval, clearance or authorization.

How long does it take to see results?

The trials disagree, and the pattern in the newest one is peculiar: WOMAC pain improved at 18 to 21 days and then no longer did at one month, while stiffness and daily activity only reached significance at one month. Objective function tests did not move at all. If you are testing a device on yourself, pick one outcome you can actually measure, give it four to six weeks of consistent use, and make sure the return window is longer than your test.

Does PEMF help you sleep?

Improved sleep is one of the most common claims in the category and one of the thinnest on evidence. A PubMed search for randomised trials of PEMF and sleep does return results, about nineteen of them, but read the list and it falls apart: post-COVID fatigue, vagus nerve stimulation through the neck, multiple sclerosis pain, cervical disc herniation, and several studies where the exposure is mobile-phone radiofrequency rather than a therapeutic device at all. Sleep is usually a secondary outcome inside a pain trial. There is no pooled analysis of PEMF for sleep that resembles the fourteen-trial low back pain review or the nine-trial knee review above. On top of that, a mat that delivers twenty minutes of heat and enforced stillness before bed has two perfectly good explanations for better sleep that have nothing to do with the magnetic field. If sleep is your goal, treat a PEMF mat as an expensive way to buy a wind-down routine.

Is PEMF the same as an MRI, a magnet, or TMS?

No, and the gaps are enormous. A static magnet does not change over time, so it induces no current at all. An MRI runs at 1.5 to 3 tesla, which is 15,000 to 30,000 gauss against a home mat's 2. Transcranial magnetic stimulation is pulsed like PEMF but strong enough to depolarise neurons directly, which is why it is a supervised clinical treatment. Sharing vocabulary with clinical technologies is not evidence of shared effects.

Is PEMF the same as red light therapy?

No. PEMF uses coils to produce a magnetic field. Red light therapy uses LEDs, typically around 660 and 850 nanometers. They are commonly sold in the same mat, which is why several products marketed as PEMF turn out to be far-infrared heating pads with no coils in them. A real PEMF device publishes a frequency range in hertz; if the only numbers on the listing are temperatures and wavelengths, it is a heating pad.

Is a PEMF device worth buying?

As an experiment on yourself with a real return window, and only if the price does not hurt, it is defensible. As a treatment, it is not something the evidence supports. The most useful spec on any of these product pages is the length of the money-back guarantee, because that is what determines whether you can find out cheaply.

Who should not use PEMF therapy?

The contraindication that matters most is an implanted electronic device. Manufacturers are consistent on this, and HealthyLine's FAQ states that it is not recommended to use PEMF therapy if you have a pacemaker or any kind of defibrillator, and to consult your healthcare provider before using any of the therapies. The same caution normally extends to insulin pumps, neurostimulators and cochlear implants, because a pulsed magnetic field can interfere with device electronics. Brands also routinely advise against use in pregnancy. If you have epilepsy, an active malignancy or a transplanted organ, the sensible move is to ask the doctor managing that condition rather than the company selling the mat. On combination mats that also produce heat, the heat carries its own separate limits and is the more likely source of a problem.

What does PEMF therapy feel like?

We have not used a PEMF device ourselves, so treat this as reasoning from the specifications rather than a report. At the intensities home mats publish, around 2 gauss, there is no obvious reason you would feel the magnetic field at all, and manufacturers do not claim you will. What you feel on most mats sold as PEMF is the other therapy in the same product. The HealthyLine TAJ series heats to 158°F, so the sensation people are describing when they say a session felt good is largely a very warm gemstone surface. This is precisely why a pure PEMF device is more useful as an experiment: if there is no heat, no light and no vibration in the session, anything you notice is at least attributable to the thing you were testing.

Our Verdict

If we were spending our own money to try PEMF today, it would be the OMI Full Body Mat at around $1,050, and it would be framed as an experiment with a deadline rather than a purchase. Pure PEMF with no heating layer is the only version of this device that can tell you anything: on a gemstone mat heated to 158°F, the heat is doing enough work that you will never know what the coils contributed. Oxford Medical publishing 210 microtesla at 8 Hz and 220 at 99 Hz, then stating in its own product sheet that this is not a medical device, is the level of candour we would want from anyone taking four figures off us.

The wider read is less comfortable. The strongest evidence sits in chronic low back pain, in trials mostly run at field strengths five to fifty times what a home mat produces. The most-cited use, knee osteoarthritis, produced effects the 2026 meta-analysis itself says fall short of clinical meaningfulness. Buy with a 90-day window, measure one thing, and be willing to send it back.