Red Light Therapy for Dogs: What the Vet Trials Show and Which Devices Publish Real Numbers

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

Red light therapy has real controlled evidence in dogs for osteoarthritis pain — and it was produced by clinician-operated therapeutic lasers, in a clinic, at 10 to 20 J/cm². That gap between the evidence and the hardware being sold to you is the whole story of this category, and almost nobody selling a pet device will point it out.

We should be straight about where we sit. We own one red light panel, mounted at standing height, and no pet device at all. Nothing here is us treating a dog. What we can do is what we do with every panel that crosses this site: read the trials, pull the manufacturers' own published specs, and check whether the numbers on the box are enough to reproduce what the researchers actually did. On this category the answer is mostly no, and which devices fail that test is genuinely useful information.

Last reviewed: August 2026

Quick comparison

Product	Best For	Price	Coverage	Wavelengths	Irradiance (mW/cm²)	Rating
Hooga HG300	Dosing you can calculate	~$199	31" × 28" at 6"	660 / 850 nm	73 published · 69 peak, 55 avg measured	4.5
Photizo Vetcare	Fixed pre-set cycles	~$397	Handheld head	633 / 850 nm	Not published	3.8
LumaSoothe 2	The best-known pet device	~$190	Handheld head	940 / 650 / 520 / 470 / 590 nm	Not published	3.5
SHINE VETLIGHT Pro	Budget handheld	~$130	Handheld head	660 / 850 nm	Not published	3.5
Clocness dog belt	Hands-free on a wriggly dog	~$60	16" × 8" pad	660 / 850 nm	">100" claimed at contact, unverified	3.3
GOVW HHL02	Closest to clinic hardware	~$208	Handheld probe	660 / 810 / 980 nm	Not published	3.4

Prices and specs read off Amazon and each brand's own site on 19 August 2026 and they move, so confirm the current price before you buy. The GOVW listing carried an $11 clip coupon that day, and the Clocness listing showed $59.99 against a $69.99 list price.

What the evidence actually says, and what it was produced with

Two randomised controlled trials carry most of the weight here, and both are worth knowing in detail because the detail is where the home devices fall down.

Looney, Huntingford, Blaeser and Mann (Canadian Veterinary Journal, 2018) ran a randomised, double-blinded, placebo-controlled multi-centre study in 20 client-owned dogs with naturally occurring elbow osteoarthritis. Eleven dogs received photobiomodulation at 10 to 20 J/cm² to both elbows over six weeks; nine received sham light at 0 J/cm². Nine of the 11 treated dogs reduced their NSAID dose. None of the nine sham dogs did. That is the single most persuasive result in the field, because a dog cannot experience placebo analgesia on its owner's behalf and the outcome was a drug dose, not an opinion.

The paper specifies its hardware, and it is worth reading closely: a 12 W gallium-aluminium-arsenide diode emitting 980 nm, run at 5 to 10 W, with a visible 3.5 mW aiming beam at 650 nm. The sham group received that aiming beam and nothing else. Five to ten watts of output is Class IV power by any standard classification, though the paper never applies the label.

Alves, Santos, Jorge and Carreira (American Journal of Veterinary Research, 2022;83(8)) went further and used an active comparator. Forty joints in 20 dogs with bilateral hip osteoarthritis were split between a 21-day course of meloxicam and photobiomodulation at 14.3 to 19.5 J/cm² delivered by a Class IV therapeutic laser over three weeks. The paper names the device — a Companion/LiteCure CTC Class IV unit — and its settings: 980 nm for dark-coated dogs, or an 80/20 blend of 980 and 808 nm for lighter coats, delivered at 4.2 to 5.2 W/cm² at the skin in a moving contact pass. The laser group came out ahead on pain severity and function at day 8, on function and gait at day 30, and improved joint range of motion from day 15 through day 90, after treatment had stopped.

Now the counterweight, which is just as real. Millis and Bergh's systematic review (Animals, 2023;13(4):667) screened the veterinary laser literature from 1980 to 2020, included 45 articles of which 24 concerned dogs, and concluded that "despite being frequently used in animals, there is no consensus regarding optimal treatment protocols nor its clinical efficacy for the treatment of various conditions." The Merck Veterinary Manual, covering the same modality, gives a working range of 4 to 50 J/cm² and states outright that "the specific dose(s) of laser required to heal tissue and treat pain remains unclear."

So: a plausible mechanism, two decent trials on arthritic joints, a wide and unsettled dose range, and clinic-grade lasers doing the delivering. Everything below is a consumer LED device trying to land somewhere inside that range.

Why irradiance is the only spec that matters, and why nobody publishes it

Dose is irradiance multiplied by time. That is the entire arithmetic. If a device tells you it emits 55 mW/cm² at six inches, you can work out that 10 J/cm² takes about three minutes and 20 J/cm² takes about six. If it tells you only that it emits "660 nm and 850 nm", you have no way to know whether your 15-minute session delivered 3 J/cm² or 40.

Here is what we found checking the pet category against its own published material on 19 August 2026:

• LumaSoothe publishes wavelengths on its own site (940 nm infrared, 650 nm red, 520 nm green, 470 nm blue, 590 nm yellow), plus the device's height, width and weight. No LED count, no mW/cm², no joules per session.
• Photizo markets the Vetcare on pre-programmed "measured dose" cycles, which is exactly the right idea. Neither the manufacturer's own product page nor its main US distributor publishes what that dose is in joules, or the output in mW.
• SHINE lists 850 nm infrared and 660 nm deep red on the VETLIGHT Pro's Amazon listing, and no output figure.
• GOVW lists its diode counts by wavelength (16 × 660 nm, 4 × 810 nm, 4 × 980 nm), then gives four power levels given as percentages of an absolute figure it never states.

Four devices, four different ways of not answering the question. Meanwhile the Hooga HG300, a plain human panel, publishes "over 73 mW/cm² irradiance at 6 inches". Unusually, that figure has been independently checked on a spectrometer by Alex Fergus at Light Therapy Insiders, who recorded 69 mW/cm² peak and 55 average. That is a 95% match on peak, and it is the reason it heads the table here. We go into the wider pattern in our Hooga review, where four other Hooga panels measured at roughly half to two-thirds of their published claims.

One honest limit on all of this arithmetic. What you can calculate is the dose arriving at the coat, at a stated distance. Fur, skin, pigment and depth all take a cut before anything reaches an arthritic hip, and we could not find a figure quantifying that loss for a home LED panel on a dog. The Merck manual notes that longer wavelengths, 810 nm and 980 nm, are the ones chosen when the target is bone or nervous tissue, which tells you the industry treats depth as a real problem rather than a solved one. Calculating your surface dose stops you under-dosing by an order of magnitude. It is not a claim about the joint.

There is a second limit, and it is the larger of the two. The hip trial did not only deliver its 14.3 to 19.5 J/cm²; it delivered them at 4.2 to 5.2 W/cm², in a moving contact pass with the head against the dog. Hooga's measured 55 mW/cm² at six inches is around a hundredth of that irradiance. Since the dose-response in photobiomodulation is biphasic and irradiance-dependent, matching the joules at a hundredth of the power density is not established to be the same treatment, and Looney's own paper points the same way: "Studies using PBMT for human chronic musculoskeletal pain have shown greater effectiveness with treatments delivering higher power densities/irradiance." Working out your dose still beats guessing at it. It is not the same thing as running the trial protocol at home, and none of the six devices below is close to being that.

The devices

Hooga HG300: dosing you can calculate

~$199 · 60 LEDs · 660 nm + 850 nm · 73 mW/cm² published at 6", 69 peak and 55 average measured · 31" × 28" coverage · 3-year warranty

Not marketed for pets, which is the reason to consider it. It is the only device in this comparison where you can put a defensible number on what a session delivered, because someone outside the company has put an instrument in front of it.

The practical shape of it is a tabletop panel on an adjustable stand. For a dog that means you position the panel, put the dog in front of it, and keep it there. No contact, no probe to hold, no fur pressed against a hot emitter. For an arthritic older dog that already lies down for long stretches, that works better than it sounds. For a springer that will not settle, it does not.

Two details from the box that matter here. It ships with protective eyewear, which is meant for you but is a useful reminder of what you are pointing at. And Hooga's separate red and near-infrared controls let you run 850 nm alone, which is the wavelength band the Merck manual associates with reaching deeper tissue.

Pros

• The only device here whose published irradiance has been independently measured
• Separate 660 nm and 850 nm controls, so you can run near-infrared alone for deeper targets
• 3-year warranty and a 60-day in-home trial, which almost nothing in the pet category matches
• ~$199 undercuts three of the four purpose-built pet devices

Cons

• Nothing about it is designed for an animal. No contour, no strap, no pet-safe eye consideration
• Requires a dog that will stay in one place for several minutes
• Coverage is a flat rectangle, so a hip gets treated at a worse angle than a chest

Best for: owners of settled, older, arthritic dogs who want to know what dose they delivered.

Check price →

Photizo Vetcare: fixed pre-set cycles

~$397 · 633 nm + 850 nm · one-touch pre-programmed dose cycles · handheld, rechargeable · model VETCARE1

The design philosophy is the right one. Instead of asking an owner to compute joules, Photizo runs a fixed cycle and stops. You press the head against the site, it delivers its dose, it beeps, you move on. For anyone who is never going to do the arithmetic, that removes the most common way home treatment goes wrong.

What we cannot check is the dose itself. Neither photizousa.com nor Miridia Technology, its main US distributor, publishes the output in mW or the energy per cycle in joules. So the device is promising a measured dose and declining to tell you the measurement. At ~$397, the same figure on Amazon and at Miridia, which lists it against a $425 regular price, that is a lot of trust to extend.

It is also the thinnest evidence base on customer side: 12 Amazon ratings at 4.4 stars as of 19 August 2026, against 1,022 for the LumaSoothe. That is not a quality signal in either direction, but it does mean there is very little owner reporting to read before you spend.

Pros

• Fixed-dose cycles remove the timing decision entirely
• 633 nm and 850 nm is a sensible pairing for surface and deeper targets
• Contoured handheld head suits joints far better than a flat panel
• Same price direct and on Amazon, so no channel-hunting needed

Cons

• Publishes no output figure anywhere we could find, so its "measured dose" is unverifiable
• Most expensive device here by a wide margin
• 12 Amazon ratings means almost no independent owner reporting

Best for: owners who want a device that decides the session length for them and will accept the manufacturer's word on what it delivers.

Check price →

LumaSoothe 2: the best-known pet device

~$190 · 940 / 650 / 520 / 470 / 590 nm across two modes · handheld · 5.6" × 3" · 4.2 stars from 1,022 Amazon ratings

The category's default recommendation, and the one with by far the most owner feedback behind it. Two modes: a deep mode combining 940 nm infrared with 650 nm red and 520 nm green, and a surface mode running 650, 520, 470 and 590 nm for skin, coat and wound work.

Five wavelengths sounds like more device for the money. It is worth understanding what the extra three are doing, which is working on the surface. Blue at 470 nm and yellow at 590 nm are not penetrating to a hip joint, and the veterinary trial evidence for joint pain was produced with near-infrared alone, at 980 nm. If your problem is a hot spot or a slow-healing graze, the surface mode is genuinely the relevant tool. If your problem is arthritis, you are paying for modes that do not address it.

The gap in the spec sheet is the same as everywhere else, and worse for being on the company's own site: LumaSoothe publishes the physical dimensions of the device and the nanometre figure for all five colours, then stops before any output figure. Independent reviewers have raised power output and fur penetration as concerns; we could not find a spectrometer measurement to settle it either way, so we are not going to pretend the question is answered.

Pros

• By far the deepest pool of owner reporting in the category, at 1,022 Amazon ratings and 4.2 stars
• Two genuinely distinct modes, with the surface mode useful for skin and coat problems
• Publishes exact nanometre figures for all five wavelengths, which most rivals do not
• Sold and shipped by the manufacturer, so warranty questions go to one place

Cons

• No irradiance, LED count or joules published anywhere on the manufacturer's own site
• Three of the five wavelengths are surface-only and irrelevant to joint pain
• Small treatment head means a large dog's hip takes several repositionings
• ~$190 buys a device you cannot dose, against ~$199 for one you can

Best for: owners dealing with skin, coat or wound problems rather than deep joint pain, who want the device most other owners have used.

Check price →

SHINE VETLIGHT Pro: budget handheld

~$130 · 660 nm + 850 nm · handheld, rechargeable · 4.4 stars from 96 Amazon ratings

The cheapest handheld here, and the one with the least padding in its spec sheet. No green, no blue, no yellow. Just deep red at 660 nm and near-infrared at 850 nm, which is the pairing the consumer red light industry has standardised on. Be precise about what that buys, though, because the marketing in this category is not. It is the mainstream human-panel pairing. It is not the pairing the dog trials used: Looney ran at 980 nm and Alves at 980 nm, or 980 blended with 808. No canine osteoarthritis trial has tested 660 or 850 nm.

That focus is worth something. It is also all the information you get: the listing publishes the two wavelengths and no output figure, so the same problem applies as everywhere else in this category. At ~$130 the stakes are lower, and if you want to find out whether your dog responds at all before committing $400, this is a defensible way to run that experiment.

Ergonomically it is the simplest thing here to hold: a shaped body you press against the site, portable enough to use wherever the dog has settled rather than making the dog come to it.

Pros

• Sticks to 660 nm and 850 nm, with no surface-only wavelengths padding the spec sheet
• Cheapest way into a purpose-built pet device at ~$130
• Rechargeable and portable, so you treat the dog where it lies
• 4.4 stars across 96 ratings is a reasonable base for the price

Cons

• No irradiance or joule figure published, so session length is guesswork
• 660 / 850 nm is the human-panel standard, not the 980 nm the canine trials ran at
• Small emitter area means slow going on a large dog
• 96 ratings is a thin base compared with LumaSoothe's 1,022

Best for: first-time buyers who want to test whether their dog responds before spending panel money.

Check price →

Clocness dog belt: hands-free on a wriggly dog

~$60 · 120 dual-chip LEDs · 660 nm + 850 nm · 16" × 8" pad · straps to 33" · 15-minute auto-off · 4.4 stars from 23 Amazon ratings

This one solves a problem the others ignore. Holding a probe against a moving dog for six minutes, several times a week, is the reason home treatment courses get abandoned. A padded wrap that straps around a hip or back and shuts itself off after 15 minutes turns a chore into something you can do while the dog sleeps.

The listing claims "contact irradiance > 100 mW/cm²". Take that as a marketing figure rather than a measurement: it is a claim about output pressed directly against the skin, from a brand whose listing first appeared in July 2025 with 23 ratings behind it, and nobody independent has put a meter on it. It is also not a figure you can line up against the rest of the table: it is a contact reading, taken with the emitters against the skin, while Hooga's 55 mW/cm² is measured at six inches. In absolute terms there is nothing far-fetched about exceeding 100 mW/cm² at contact — Merck puts usual clinical power densities at 10 to 100 mW/cm², and the Alves trial laser delivered 4,200 to 5,200 mW/cm² at the skin. The problem is not that the number is too big. It is that it is unverified, and measured on a different axis to every other number in the comparison.

What it is unambiguously good at is contact and coverage. A 16 by 8 inch pad wrapped against a hip puts the emitters at zero distance across a large area, which is a fundamentally better geometry than a flat panel two feet away, whatever the absolute output turns out to be.

Pros

• Hands-free, which is the single biggest practical barrier to finishing a course
• 16" × 8" contact area beats every handheld here for coverage per session
• 15-minute auto-off removes the risk of leaving it running
• ~$60 makes it cheap enough to combine with something else

Cons

• The ">100 mW/cm²" claim is unverified, and a contact reading cannot be compared with an at-distance figure
• 23 ratings and a July 2025 listing date means almost no track record
• USB-powered pad construction is the least durable format here
• Straps fit a range of dogs but nothing about the geometry is dog-specific

Best for: owners of dogs that will not sit still, and anyone who has already abandoned one treatment course.

Check price →

GOVW HHL02: closest to the clinic hardware

~$208 · 16 × 660 nm, 4 × 810 nm, 4 × 980 nm · four power levels · continuous and pulse modes · 1–60 min timer · 4.2 stars from 183 Amazon ratings

The only device here using 810 nm and 980 nm, which are the wavelengths the Merck manual identifies as passing through superficial layers to reach bone and nervous tissue. It is also the only device in this roundup whose wavelengths actually match the trials. 980 nm is exactly what Looney's elbow study emitted, and 810 nm sits 2 nm away from the 808 nm Alves blended in for lighter-coated dogs. No other device here emits at either wavelength, our pick included. The rest of the field runs the 660 and 850 nm pairing that human panels standardised on, or in LumaSoothe's case a 940 nm mix, and none of those has a canine osteoarthritis trial behind it.

It is also a laser device rather than an LED array, and that changes the safety picture. Retinal exposure is the one absolute contraindication in veterinary photomedicine, and a collimated laser beam is far less forgiving of a badly aimed session than a diffuse panel. Neither the listing nor the packaging we could see states a laser class or an absolute power figure. The four power levels are given as 25%, 50%, 75% and 100% of a number that is never printed. One Amazon reviewer describes it as a 1.7 W Class 3B unit; that is one person's reading, not a manufacturer spec, and we are not going to treat it as one.

Its own listing recommends three to five sessions a week for four weeks at 15 minutes each, which is at least a concrete protocol and broadly consistent with the trial courses.

Pros

• The only device here whose wavelengths match the trials: 980 nm as in Looney, 810 nm against Alves's 808 nm
• Adjustable power, continuous and pulse modes, and a 1–60 minute timer
• Publishes diode counts per wavelength, which most rivals do not
• Concrete manufacturer protocol rather than "use as needed"

Cons

• Power levels are percentages of an absolute figure the manufacturer never states
• Laser rather than LED, so aiming errors carry more eye risk
• No stated laser class anywhere we could find on the listing
• Small probe area makes a large dog a long session

Best for: owners already working with a vet on laser therapy who want to continue between clinic visits, with the vet's guidance on aiming and dose.

Check price →

Buyer's guide

Start with the diagnosis, not the device

Both trials that make this category interesting enrolled dogs with a confirmed osteoarthritis diagnosis. The Merck manual's contraindication list includes areas of neoplasia or malignancy, and lumps get mistaken for arthritis routinely. Treating an undiagnosed mass with a modality whose entire mechanism is stimulating cellular activity is the one genuinely bad outcome available here. Get the limp diagnosed first.

Decide whether you are dosing or trusting

There are only two coherent positions. Either you buy a device that publishes a verified irradiance figure and you calculate your sessions, or you buy a device with fixed pre-set cycles and accept the manufacturer's judgement. What does not work is buying an adjustable device with no published output and guessing. That is the combination that produces a 15-minute session delivering an unknown fraction of a therapeutic dose.

Match the format to the dog, not to the spec sheet

The best device is the one you will still be using in week four. A panel needs a dog that settles. A handheld needs an owner with the patience to work a hip in sections. A wrap needs neither, which is why it is worth considering even with the weakest spec sheet in the comparison. Adherence beats irradiance if the alternative is an abandoned course.

Take the eye risk seriously, especially with lasers

Retinal damage from direct or reflected exposure is the one thing the veterinary sources treat as absolute rather than precautionary. It applies more sharply to the collimated laser devices than to diffuse panels, but it applies to both. Angle the emitter away from the head, do not treat around the face without your vet's guidance, and do not let a curious dog turn round into the beam.

Do not expect the regulatory framework to filter anything

The FDA does not require a 510(k), a PMA or any pre-market approval for devices intended for animal use, though it keeps oversight and can act if an animal device is misbranded or adulterated. In practice that means no agency has reviewed the output claims on any of these listings. It is also why we lean so hard on the one figure with an independent measurement behind it. Our guide to FDA-cleared red light therapy devices covers what clearance does and does not mean on the human side.

FAQ

Does red light therapy actually work for dogs?

For osteoarthritis pain, there is genuine controlled evidence. Looney and colleagues (Can Vet J 2018) randomised 20 dogs with naturally occurring elbow osteoarthritis to photobiomodulation or sham light for six weeks: 9 of 11 dogs in the treated group reduced their NSAID dose against 0 of 9 in the sham group. Alves and colleagues (Am J Vet Res 2022) compared photobiomodulation against meloxicam across 40 hip joints and found benefits for pain, function and gait that outlasted the treatment period. The caveat that matters is what delivered it: both trials used clinician-operated therapeutic lasers in a clinic at 10–20 J/cm². The hip trial names a Companion/LiteCure CTC Class IV laser running at 980 nm, or an 80/20 blend of 980 and 808 nm for lighter-coated dogs; the elbow trial specifies a 12 W gallium-aluminium-arsenide diode at 980 nm delivered at 5 to 10 W, which is Class IV territory by power even though the paper never uses the label. No consumer LED device has been through a trial like that, so the honest position is that the mechanism has support in dogs and the home hardware does not.

Do vets recommend red light therapy for dogs?

Many veterinary rehabilitation practices offer it, and the Merck Veterinary Manual covers it as an accepted modality with energy densities ranging from 4 to 50 J/cm². It also says plainly that "the specific dose(s) of laser required to heal tissue and treat pain remains unclear". Millis and Bergh's systematic review (Animals, 2023) screened literature from 1980 to 2020, included 45 articles of which 24 concerned dogs, and concluded that "despite being frequently used in animals, there is no consensus regarding optimal treatment protocols nor its clinical efficacy". Treat it as a reasonable adjunct your vet may support, not a replacement for a diagnosis or for pain medication.

How often should I use red light therapy on my dog?

The two controlled trials give you a defensible starting pattern. Looney's elbow study treated both elbows over six weeks; the hip study ran a three-week course and measured effects out to 90 days. Manufacturer guidance for home devices lands in similar territory. The GOVW listing recommends three to five sessions a week for four weeks at 15 minutes each. A course of several weeks, then a break to see whether anything changed, is more informative than treating indefinitely. Ask your vet first if your dog is on medication, because the point of the elbow trial was that the treated dogs needed less of it.

How long should a red light therapy session be for a dog?

It depends entirely on the device's output, which is why irradiance matters more than any timer preset. Take the Hooga HG300, measured independently at 55 mW/cm² average at six inches. Dose is irradiance multiplied by time, so 10 J/cm² takes roughly three minutes at that distance and 20 J/cm² takes about six. Work the same sum with the panel's published 73 mW/cm² instead and you get about two and a quarter minutes for 10 J/cm², a third short. That is surface dose at the coat, not dose at the joint; fur, skin and depth all take a cut that nobody has quantified for a home LED panel. Our guide to how long a red light therapy session should be works through the same arithmetic for human use.

Can a dog have too much red light therapy?

Overdosing is a real concept in photobiomodulation, because the dose-response is biphasic, meaning more is not linearly better. On harms rather than dose, the Merck Veterinary Manual names retinal damage and thermal burns from improper application as the main risks of laser therapy. Heat is the one a home device makes easy. An LED panel held close for a long session warms the skin, and a dog cannot tell you when it is uncomfortable. Keep sessions to the low single-digit minutes per site at panel distances, and stop if your dog moves away.

What are the side effects of red light therapy for dogs?

The Merck Veterinary Manual puts the main risks of laser therapy at retinal damage and thermal burns when it is improperly applied. The eye is where it is bluntest: laser light can damage the retina "whether reflected off shiny surfaces or shown directly into the eye". Among the contraindications Merck lists are areas of neoplasia or malignancy, a hyperactive thyroid gland, areas of active haemorrhage, the retina and a pregnant uterus. The list continues past those, and two of the rest matter here: laser therapy is also contraindicated for patients with lymphoma or on immunosuppressant medications, and caution is warranted over the long bones of animals under a year old, because higher-powered devices may stimulate premature closure of the growth plates. Do not treat over an undiagnosed lump, and do not let your dog look into the emitter. Our red light therapy side effects guide covers the human picture in more depth.

Can I use my own red light panel on my dog?

Yes, and on the published numbers it is the more transparent option. Every dedicated pet device we checked publishes its wavelengths and stops there. LumaSoothe's own site lists 940, 650, 520, 470 and 590 nm and no irradiance, LED count or joules; Photizo's manufacturer pages publish neither. A human panel that states mW/cm² at a stated distance lets you calculate the dose, which is the only number the trials were controlling. The trade-offs are that a panel is not handheld, it is not contoured to a joint, and you have to keep a dog still in front of it. Our red light therapy wavelengths explainer covers which bands do what.

What is the best red light therapy device for dogs?

If you want to reproduce a known dose, a human panel with a verified irradiance figure. The Hooga HG300 at ~$199 is the cheapest we know of that has been independently measured. If you want something handheld and contoured to a joint, the Photizo Vetcare at ~$397 runs fixed pre-programmed cycles at 633 and 850 nm, which removes the timing decision, though it publishes no output figure to check that dose against. If you mainly need hands-free, a strap-on pad like the Clocness dog belt at ~$60 solves the practical problem of holding a device against a moving dog, at the cost of an irradiance claim nobody has verified.

Our Verdict

If we were buying today for an arthritic dog, we would buy the Hooga HG300 at ~$199. Not because it is designed for animals, because it is not, but because it is the only device in this comparison where we could tell you what a session delivered and show our working. An independently measured 55 mW/cm² average puts 10 J/cm² at about three minutes and 20 J/cm² at about six, which brackets the range both trials used. Every purpose-built pet device we checked asks you to accept a dose it will not print.

The honest second pick is the Clocness belt at ~$60, bought alongside rather than instead. Its irradiance claim is unverified and we would not build a dose on it, but it solves the problem that actually ends most home treatment courses: a dog that will not hold still for six minutes. A course you finish with an uncertain dose beats a course you abandon with a calculated one.

And before either: get the limp diagnosed. The evidence here is for osteoarthritis, and the one contraindication everybody agrees on is treating something you have not identified.

More in our red light therapy section, and you can read about who we are and what we actually own.