Average Power in Red Light Therapy: What It Means, Why It Matters, and How to Evaluate a Device

Deep dive

The science: what average power measures

The essential terms: power, irradiance, energy, and dose

PBM product pages often use the word “power” to refer to several different measurements. They are related, but they are not interchangeable.

Term	Unit	What it tells you	What it does not tell you
Average optical power	W or mW	Average rate of optical energy output	How concentrated the light is or how much reaches a deep target
Peak optical power	W or mW	Highest instantaneous power during a pulse	Total energy emitted during treatment
Pulse energy	J or mJ	Energy in one pulse	Total session energy without pulse frequency and time
Irradiance	W/cm² or mW/cm²	Optical power arriving per unit area	Total energy without treatment time
Radiant energy	J	Total optical energy emitted or delivered	How that energy is distributed across an area
Radiant exposure	J/cm²	Energy delivered per unit area	How quickly it was delivered or how much reached depth
Electrical input power	W	Electricity consumed by the device	Therapeutic optical output

Researchers have warned for years that missing parameters, terminology errors, and unverified output make PBM studies difficult or impossible to reproduce. Jenkins and Carroll identified wavelength, power, time, beam area at the treatment surface, and pulse parameters among the minimum information needed for a repeatable study. Hadis and colleagues reviewed 74 PBM papers and found that 41% did not report power, 40% did not report beam area, 43% did not report irradiance, and 73% did not explain how the light was measured. A 2025 proposed reporting framework noted that missing data and unverified specifications remain problems in laser research (Jenkins and Carroll, 2011; Hadis et al., 2016; Shurrab, 2025).

These aren’t just paperwork slip-ups. If the power or area is off, every calculation that depends on them is off too.

The core power calculation

Power is energy per unit time:

Energy (J) = average power (W) x treatment time (s)

Examples:

Average power	Time	Energy
0.05 W	60 s	3 J
0.10 W	60 s	6 J
0.50 W	60 s	30 J
1.00 W	60 s	60 J
1.00 W	30 s	30 J

This math answers a simple but crucial question: how much optical energy did the device actually deliver across the measured area during the session?

But it doesn’t tell you how that energy is spread out. A 1 W probe over 1 cm² and a 1 W pad over 100 cm² both put out the same total power, but their average irradiance is worlds apart.

How average power connects to irradiance

If optical power is distributed uniformly across a known area, average irradiance can be estimated as:

Average irradiance (W/cm²) = average power (W) / treatment area (cm²)

Radiant exposure can then be calculated as:

Radiant exposure (J/cm²) = average irradiance (W/cm²) x time (s)

Notice what both formulas start with: average power. If that number is missing, or if it’s really a peak value, everything calculated from it is wrong.

Average optical power is not electrical wattage

You might see wattage listed for the power adapter, battery, LED chips, or just electrical use. None of those numbers tell you how much optical power the device actually puts out.

Electrical energy is divided among several paths:

• Light output at different wavelengths
• Heat generated in the emitters
• Driver circuitry
• Control electronics
• Fans, displays, and other components
• Conversion losses

A 20 W adapter does not mean the device gives 20 W of optical power. Similarly, 100 LEDs rated at 1 W each do not add up to 100 W of light output. LEDs may run below their rating; they aren’t 100% efficient, and some light may not reach the treatment area.

Look for wording such as:

Measured average optical output power: 1.1 W at the treatment surface during the standard operating mode.

Be careful with vague labels like “device power,” “total wattage,” “maximum power,” or “100 W LED system.” These are incomplete unless the manufacturer explains exactly what was measured.

Average power versus peak power

Peak power is the maximum light energy produced at any single moment within a pulse. In contrast, average power is calculated over a set period of time, factoring in both the times when the light is on and when it is off or emitting less.

For a stable continuous-wave source:

Average power is approximately equal to operating power.

For an ideal rectangular pulsed source:

Average power = peak power x duty cycle

Duty cycle is the fraction of time the source is on. A 10 W pulse operating at a 1% duty cycle has an average power of:

10 W x 0.01 = 0.10 W

Over 60 seconds, its total energy is:

0.10 W x 60 s = 6 J

Reporting only “10 W peak power” would make the output appear 100 times greater than the average power used to calculate session energy. Plugging that peak number into the irradiance formula would make the dose you calculate 100 times too high.

Here’s what that looks like with a second example, a 10 W peak at a 5% duty cycle:

Figure 2. A pulsed device spends most of its time off. The peaks reach 10 W, but the average, the number that goes into irradiance and dose, is only 0.5 W. Chart: MedcoVet.

Figure 3. Average power is the steady level that delivers the same energy as the pulses. The two shaded areas are equal. Chart: MedcoVet.

For a pulse train, average power can also be calculated from pulse energy:

Average power = energy per pulse x pulse frequency

The basic peak-power calculation is accurate only if you know the pulse shape and duty cycle precisely. In practice, pulses can vary in height and duration, or include complex waveforms. Therefore, measuring the average power over time gives a more dependable result than relying on theoretical calculations.

Does peak power matter?

Yes, but peak power tells you something different.

Peak power, pulse duration, frequency, duty cycle, and energy per pulse describe how the exposure is arranged in time. Pulse structure can affect thermal behavior and may affect biological response. A review of pulsed PBM found examples favoring pulsed delivery, examples favoring continuous delivery, and many conditions where evidence was insufficient to choose one mode (Hashmi et al., 2010). Pulse parameters were also reported poorly and inconsistently.

In a mouse traumatic brain injury study, researchers delivered continuous 810 nm light and 810 nm light pulsed at 10 Hz or 100 Hz with the same average irradiance and total fluence. Outcomes differed, with the 10 Hz group showing better recovery on the reported measure at 28 days (Ando et al., 2011). This animal study used one injury model. It does not identify a universal pulse frequency, but it shows why average power and total energy do not fully describe a pulsed exposure.

In another experiment, a 200 mW continuous 810 nm laser and a 60 mW superpulsed 904 nm laser showed different transmission profiles through isolated rat skin (Joensen et al., 2012). The devices differed in wavelength, power, and pulse structure, so the result cannot prove that pulsing alone caused the difference. It does show that you can’t interpret a peak-power claim without the other device parameters.

The correct conclusion is:

• Peak power is relevant to pulse behavior.
• Average power is required to calculate total energy over time.
• Neither number proves clinical effectiveness on its own.
• A pulsed device should disclose both, along with the complete pulse settings.

Continuous, pulsed, and superpulsed output

Mode	What it means	What the company should report
Continuous wave	Stays on at a relatively steady level	Measured average power and output stability
Gated or pulsed	Switches on and off at a stated pattern	Average power, peak power, pulse width, frequency, duty cycle, and pulse energy
Superpulsed	Produces very short pulses that may have high peaks but modest average output	The same pulse data, plus measured average power in the treatment mode

Superpulsed just tells you how the device operates, not how well it works. A device can have sky-high peak power and still deliver low average output.

Why a high total wattage can still produce low irradiance

Total power must be read with the treatment area.

Consider two simplified devices:

Device	Average power	Treatment area	Average irradiance
Small probe	1 W	2 cm²	500 mW/cm²
Large pad	5 W	100 cm²	50 mW/cm²

The pad has five times the total power, but the probe has ten times the area-average irradiance. Neither comparison says which device is appropriate. The probe may treat a small point quickly and may require careful thermal control. The pad may cover a broad area more evenly and take longer. That’s why you can’t just compare devices of different sizes by total wattage. Average power tells you the total output rate. Irradiance tells you how concentrated that output is at the treatment surface.

Average power is not the dose

The word “dose” is often used loosely in PBM. It may refer to total energy, energy per area, energy at the surface, estimated energy at depth, or an entire treatment schedule. This ambiguity has long been identified as a barrier to comparing PBM protocols (Enwemeka, 2009).

Average power is one input to dose calculation, and everything else is built on it. A complete exposure description includes at least:

• Average power
• Treatment time
• Treatment area
• Irradiance
• Radiant exposure
• Wavelength
• Pulse mode
• Number of treatment points
• Number and spacing of sessions
• Placement and contact

Two devices can emit the same total energy and still deliver very different exposures. For example, 0.5 W for 60 seconds and 1 W for 30 seconds both equal 30 J. The second device delivers that energy twice as quickly. If the treatment areas are equal, it also has twice the irradiance.

The biological effects will differ. PBM doesn’t reliably follow reciprocity, the simple rule that swapping power and time gives the same result.

What tissue responds to

Tissue does not respond to a wattage label. A photobiological response begins when molecules in cells and tissue absorb photons of suitable wavelengths. Proposed PBM mechanisms include changes in mitochondrial activity, nitric oxide signaling, reactive oxygen species, calcium signaling, gene expression, and inflammatory pathways. Active pathways can vary by wavelength, cell state, and exposure conditions (de Freitas and Hamblin, 2016).

Average power matters because it is the source term for the exposure, but the tissue experiences something more specific:

1. The device emits optical power.
2. That power spreads over an area, producing irradiance.
3. Skin, fur, blood, fat, and other tissue reduce and redistribute light.
4. A fraction reaches the target tissue.
5. Relevant molecules absorb a fraction of the arriving light.
6. The absorbed exposure can initiate a biological response.

That’s why saying average power drives healing is way too broad. Average power tells you what’s available to deliver. But it’s irradiance, wavelength, tissue transmission, absorption, time, and the biological state that decide what actually reaches the target tissue.

That’s also why two products with the same wattage can act totally differently. One might spread light over a bigger area. Another might lose more light before it even hits the skin. Some split output across different wavelengths. Some have a hot spot in the middle and weak edges. The number on the label is just the start.

Why the same joules delivered at different rates don’t give the same result (reciprocity) is explained in the Irradiance guide.

Why more average power is not always better

Higher average power can speed up treatments by delivering the same energy in less time. That can help, and it might make it easier to keep the device in place since the session is shorter. A 2026 network meta-analysis of 22 randomized trials (1,353 patients) found that high-intensity lasers reduced pain somewhat more than low-level lasers in some comparisons, but the differences were “generally modest and frequently below” the smallest change patients would notice (de la Barra Ortiz et al., 2026).

But increasing power can also:

• Raise irradiance when area stays constant
• Increase skin temperature
• Make timing errors more important
• Increase the effect of beam hot spots
• Deliver an excessive surface exposure before enough light reaches a deep target
• Move the exposure outside the range used in supporting research

PBM often has a biphasic response. Too little light may not help, and more light can help up to a certain point. After that, more light may not help or could even reduce the effect. Studies have found these patterns in several PBM models (Huang et al., 2011; Sharma et al., 2011). This does not create one universal maximum power or irradiance. The useful range changes with wavelength, tissue, outcome, treatment area, time, and schedule. Claims such as “all PBM should stay below 300 mW/cm²” or “4 to 8 J/cm² works for most conditions” are too broad to serve as general rules. The first is an irradiance claim, not an average-power claim, and the second ignores the target tissue and protocol.

Does more power make light go deeper?

Power and light penetration are related, but marketing often makes this connection seem simpler than it is. Increasing the device’s power at a constant wavelength and setup increases the total amount of light reaching deeper tissues, but it doesn’t change how much of that light tissue actually lets through. For example, if only 1% of light passes through a tissue, doubling the surface power roughly doubles the amount at depth, but the penetration percentage stays the same. The device must still deal with the other 99% of light, which is mostly absorbed or scattered near the surface and can lead to extra heat in those upper layers.

The percentage that gets through is set mostly by wavelength and the tissue in the way. That’s why wavelength, not power, is the first thing to check for reaching a deep target.

Double the power, up to about 10% deeper

So how much deeper does extra power get you? Not much. Physicist Lars Hode puts it simply: twice the power does not mean twice as deep, but maybe 5 to 10% deeper (Hode).

His example: a 100 mW laser reaches a certain intensity at 10 mm. Drive the same laser at 200 mW, and that intensity shows up at 10.5 to 11 mm. At 400 mW, it moves to 11 to 12 mm (Hode).

Figure 4. Doubling a laser’s power moves a given light intensity only about half a millimeter to a millimeter deeper. Example from Hode, Penetration of Light into Living Tissue. Chart: MedcoVet.

Why so little? Light fades fast in tissue. Every few millimeters, scattering and absorption take away a large share of what’s left. Doubling the power only buys back one of those steps. So at the same wavelength and setup, a 2 W device isn’t reaching twice as deep as a 1 W device. It reaches roughly a millimeter deeper while putting twice as much light into the surface layers, where most of it is absorbed.

Compare that with wavelength. In Hode’s analysis of light passing through a human hand, 810 nm light got through about 25 times better than 980 nm light. Contact and pressure matter too: pressing the probe into the skin pushes blood out of the light’s path, and skin pigment can absorb much of the incoming light (Hode).

What about high peak power?

Peak power tells a different, more complicated story. Lasers that fire very short, very intense pulses (superpulsing, typical of 904 nm lasers) appear to get through skin more easily. A systematic review assumed in its dose model that the skin removes about 80% of the energy from continuous 820 nm and Nd:YAG (1060 nm) lasers, but only about 50% from pulsed 904 nm GaAs lasers (Bjordal et al., 2003). In a rat skin study, the percentage of light that got through stayed at about 20% with a continuous 810 nm laser, but rose from 38% to 58% over 150 seconds with a superpulsed 904 nm laser (Joensen et al., 2012). One proposed explanation is that the intense pulses temporarily “bleach” the tissue, making it more transparent (Hode).

Two cautions before you go shopping for peak power:

• A higher percentage isn’t “more light”. In that rat study, the continuous laser put out 200 mW and the superpulsed laser 60 mW (average). Twenty percent of 200 mW is about 40 mW through the skin. Even 58% of 60 mW is only about 35 mW. The continuous laser still delivered at least as much light. (This is our rough comparison of absolute output. The authors concluded that the superpulsed laser penetrated two to three times more effectively, and the two lasers also differed in probe design and skin compression.)
• Peak power doesn’t change the dose math. The energy delivered from a pulsed laser is always the average power multiplied by time (Tunér, Hode, and Jenkins). And simple pulsing, which just switches a continuous beam on and off, doesn’t improve penetration (Hode). It only lowers the average power.

The takeaway: more power, peak or average, buys a little extra depth. Wavelength, contact, and pressure buy a lot more. And average power is still the number you need to calculate the dose.

Depth also depends on:

• Wavelength and spectral bandwidth
• Beam diameter and divergence
• Tissue water, blood, fat, and pigment
• Skin and coat characteristics
• Contact angle and pressure
• Continuous or pulsed operation

In healthy dogs, an in vivo transmission study found that power, wavelength, coat color, and shaving all affected the light measured through tissue (Hochman-Elam et al., 2020). An equine skin study found that the lower-average-power device tested transmitted more energy than the Class IV device used for comparison (Luna et al., 2020). This does not mean lower power always penetrates farther. It means total wattage and laser class did not predict transmission across two devices with different wavelengths, pulse designs, applicators, and operating methods.

Average power is just one piece of the puzzle for getting light through tissue. You still need to know how the tissue itself affects the light.

Why contact and fur change the power that reaches a pet

Average output measured in open air is not necessarily the output that reaches the skin during treatment.

How fur, distance, and contact cut the light that reaches the skin is covered in the Target Tissue guide. The WAIT Test™ counts only 25% of a device’s light if it doesn’t part the fur, and that reduced power is what A4 uses.

What veterinary research tells us about power

Veterinary PBM research supports careful parameter reporting, but it does not identify one ideal average power for every animal or condition.

A 2023 systematic review included 45 veterinary laser-therapy studies: 24 in dogs, one in cats, and the remainder in horses. Results conflicted, clinical application was often unclear, and the authors linked some of the uncertainty to wide variation in wavelength, laser class, dose, frequency, and treatment duration. Many studies also carried moderate or high risk of bias (Millis and Bergh, 2023).

Some individual studies show why complete dosing information matters:

• A randomized, blinded, placebo-controlled study in dogs with elbow osteoarthritis used 10 to 20 J/cm² per joint twice weekly for six weeks and reported improvements in lameness and pain measures (Looney et al., 2018). The result supports that this specific protocol worked better than sham in a small trial.
• The healthy-dog transmission study showed that power, wavelength, coat color, and shaving each independently affected transmission (Hochman-Elam et al., 2020).
• The equine skin comparison found that the tested device with lower average power transmitted more energy than the tested higher-power Class IV device, showing that device architecture and delivery method can outweigh total wattage in a head-to-head comparison (Luna et al., 2020).

The straightforward conclusion is that average power is necessary for evaluating a protocol, but it is not independent proof of efficacy. Evidence must match the animal, condition, target tissue, device, and treatment settings.

Technical reference

Technical details

Average power and thermal safety

PBM is supposed to work by light, not by heat. But not every device or setting avoids heating up the tissue.

Temperature rise depends on more than total power:

• Irradiance
• Wavelength
• Exposure time
• Spot size and beam profile
• Tissue absorption
• Skin and coat pigmentation
• Contact and pressure
• Whether the applicator is stationary or moving
• Cooling and heat dissipation

An in vitro investigation of high-energy PBM settings recorded clinically relevant temperature increases under several tested conditions and found that wavelength, applicator size, and movement affected heating (Cronshaw et al., 2023). The study does not set a universal safety threshold for pets or at-home devices. It supports measuring temperature under actual operating conditions rather than assuming a PBM label guarantees non-thermal delivery.

A good at-home device should make it tough to mess up the protocol. Look for features like these:

• Fixed or bounded treatment times
• Stable contact geometry
• Output limits appropriate to the treatment area
• Temperature testing under worst-case normal use
• Clear instructions for skin, coat, and wound use
• Automatic shutoff
• Eye-safety instructions suited to the source

If a pet shows discomfort or the treatment area becomes hot, stop treatment and follow the device instructions or veterinary guidance.

Why measured average power matters

A spec is only useful if it tells you what the device actually puts out.

Hadis and colleagues explain why radiometry for LEDs and lasers requires suitable, calibrated detectors; attention to wavelength response; complete beam capture; geometry; and clear reporting of the measurement method (Hadis et al., 2016). A detector can under- or over-read if the source is too large, the angle is wrong, part of the beam misses the sensor, or the instrument isn’t appropriate for the wavelength and power range.

A common example is the inexpensive “solar power meter.” These meters are calibrated for sunlight, and a detector’s response changes with wavelength, so they can misread narrow-band red and near-infrared light (Hadis et al., 2016). In one light-panel company’s comparison against a reference lab, a solar meter read three panels 31% to 88% too high (GembaRed).

In 2022, Girasol and colleagues measured 24 PBM laser devices in clinical use. Only nine produced average power within 90% to 110% of the manufacturer’s value. Across all devices, measured power ranged from 2% to 134% of the stated value. Beam diameters also differed substantially from the nominal values, and some devices changed output between one and three minutes (Girasol et al., 2022).

That research looked at older clinical laser devices, not newer consumer LED models, so the reported failure rates may not apply to all products. The important point is that the numbers listed on a device’s label do not always match what is actually measured in practice.

What a useful measurement statement should include

A manufacturer should be able to state:

• What was measured: optical power, not electrical input.
• Where it was measured: at the aperture, at a stated distance, or at the skin-contact plane.
• Which mode was measured: continuous, pulsed, red only, near-infrared only, or combined.
• How long the output was sampled: long enough to represent the treatment session.
• Which instrument was used: a detector suitable for the wavelengths and power range.
• Calibration status: when and how the instrument was calibrated.
• How the full beam was captured: especially for large pads, clusters, or divergent LEDs.
• Expected tolerance: unit-to-unit and operating variation.

Independent testing adds confidence, but a named laboratory alone is not enough. The method and operating conditions still matter.

Output stability across the full session

“Maximum power” at startup may not represent average power during treatment.

LED and laser output may fluctuate because of component warming, battery drain, temperature regulation by electronics, or shifts in pulsed program phases. To be accurate, measure average power over the actual treatment session or another well-defined relevant period.

Questions to ask include:

• Is the stated value an initial reading, a maximum reading, or a time average?
• Was the device tested from a fully and partly discharged battery?
• Does output fall when the device warms?
• Do different intensity modes have separate measured values?
• Is the same average power maintained for the full recommended session?
• Does automatic thermal control reduce output, and if so, by how much?

This is important when treatment time is based on the stated power. If output drops by 30% during a session, the energy delivered will be less than you would expect based on the starting value.

What about devices with more than one wavelength?

A multi-wavelength device should disclose how its average output is divided.

Suppose three devices all claim 1 W of total average power at 660 and 850 nm:

Device	660 nm output	850 nm output	Total
A	0.5 W	0.5 W	1.0 W
B	0.8 W	0.2 W	1.0 W
C	Not disclosed	Not disclosed	1.0 W

The three labels do not describe the same information. Devices A and B distribute optical energy differently. Device C prevents the reader from calculating energy at either wavelength.

Ask:

• Do both wavelengths operate at the same time?
• Is the stated power combined or per wavelength?
• What is the measured average power of each wavelength channel?
• Does selecting one wavelength change total output?
• Does pulse mode apply to both wavelengths in the same way?
• Were the channels measured separately with a wavelength-appropriate detector?

For an LED array, also ask whether the measured total includes the full treatment area or is an extrapolation from one emitter.

Laser class is about eye safety, not power or effectiveness

Many websites and product pages treat laser class as a power rating: for continuous-wave lasers, roughly up to 500 mW is Class 3B and over 500 mW is Class 4, and a higher class means a stronger, deeper-reaching, better device. Many also suggest that a device’s eye safety depends on its total power.

Neither is true.

Laser classification is a hazard rating. It describes how much laser light a person could get into their eye and, for Class 3B and 4, onto their skin, from the device (FDA). As Tunér, Hode, and Jenkins put it, laser classification “has nothing at all to do with suitability for laser treatment” (Tunér, Hode, and Jenkins).

What actually decides the class

The international standard, IEC 60825-1, classifies a laser product by the light that can actually reach the eye during use, not by the device’s total output (IEC 60825-1:2014). That depends on:

• Wavelength. Near-infrared light is invisible, so it doesn’t trigger the blink reflex that helps protect the eye from bright visible light.
• Power per emitter that can enter the eye. The test measures light through an opening about the size of a fully open pupil (7 mm), not the device’s combined output.
• Beam divergence and apparent source size. A tight, parallel beam can focus to a tiny, intense spot on the retina. Light that spreads out, or comes from a larger apparent source, is far less concentrated.
• Number and spacing of emitters. Emitters spaced far enough apart can’t all send light into the same pupil at once.
• Exposure time and pulsing.
• Enclosure. Light that can’t get out can’t reach an eye.

Source size alone makes a big difference. Under the standard’s limits for 808 nm near-infrared light, a point-like laser beam can deliver only about 0.6 mW through that pupil-sized opening and still be Class 1. A source that looks large to the eye can deliver about 42 mW through the same opening and still be Class 1. That’s roughly 65 times more light at the same wavelength, in the same class (IEC 60825-1:2014).

Same power, different class

Tunér, Hode, and Jenkins give a clear example. The 500 mW line people quote is the upper limit for Class 3B, so a single continuous laser just over 500 mW is Class 4. Now picture a probe with 10 laser diodes of 450 mW each, placed some distance apart. Each diode is under the limit, so the instrument is Class 3B, even though its total output is 4.5 W (the final class depends on formal testing) (Tunér, Hode, and Jenkins).

Device	Emitters	Power per emitter	Total output	Typical class
A: single laser	1	Just over 500 mW	About 0.5 W	Class 4 (over the 500 mW Class 3B limit)
B: ten-laser probe	10, spaced apart	450 mW	4.5 W	Class 3B (each emitter under 500 mW)

Figure 5. Device B puts out about nine times more total power than Device A but carries the lower class, because class depends on how much light can reach the eye from one emitter, not on the device’s total output. Example from Tunér, Hode, and Jenkins; the final class depends on formal testing. Diagram: MedcoVet.

Device B puts out about nine times more power than Device A, yet carries the lower class. As Tunér, Hode, and Jenkins put it, high power doesn’t mean a laser instrument has to be Class 4.

One caution: a lower class isn’t the same as eye-safe. The FDA describes Class 3B lasers as an immediate eye hazard when viewed directly (FDA). The point isn’t that Device B is safe to look into. The class number doesn’t tell you how much power a device delivers.

Class 1 devices can out-power Class 4 lasers

It goes further than that. “Actually, there are Class I lasers that are higher powered than many Class IV instruments!” (Tunér, Hode, and Jenkins).

That isn’t a loophole. Many Class 1 products, such as enclosed industrial laser systems, contain a Class 3B or Class 4 laser inside a housing that keeps the beam from reaching anyone’s eyes during normal use (CSUS Laser Safety Manual). And a therapy device that spreads its light widely, so only a small fraction can enter the eye at once, can carry a low class despite a high total output.

LED devices are a separate case. LEDs are no longer classified under the laser standard. Their eye and skin safety is assessed under a different standard for lamps and LEDs (IEC 62471), which uses risk groups instead of laser classes.

Class doesn’t predict effectiveness or penetration either

There is no such thing as “Class IV technology.” Choosing an emitter above 500 mW is not a different technology. Apart from power, the only differences between Class 3B and Class 4 lasers are the potential hazards and, usually, the price (Tunér, Hode, and Jenkins).

Higher class doesn’t mean deeper penetration, either. Tunér, Hode, and Jenkins point out that a high-power Class 4 laser at 980 nm can penetrate less than a 500 mW Class 3B laser at 808 nm, because water absorbs more of the 980 nm light near the surface. The extra power mostly becomes heat (Tunér, Hode, and Jenkins). A lower-power device held in contact with slight pressure can get more of its light into the tissue (Tunér, Hode, and Jenkins).

A Class 4 laser can have high average output. Still, the class label does not tell you the wavelength match, beam area, irradiance, pulse structure, treatment time, tissue transmission, or evidence for a condition. Check the real optical measurements, and follow the eye-safety instructions for the specific device. Don’t assume a higher class, bigger wattage, or the word “professional” means better PBM, and don’t assume a lower class means lower power.

That’s also why The WAIT Test™ uses class only as a safety gate, never as a score. A device sold for home use must be eye-safe without goggles. Class 3B and 4 lasers and LEDs in Risk Group 2 or 3 under IEC 62471 are marked Not safe for home use and aren’t scored. A laser with no published class needs more information until its class is published. Devices that clear the gate are then judged on their measured output, like every other device.

Does average power mean the same thing for lasers and LEDs?

The unit is the same. One watt of measured optical power is one joule per second whether the source is a laser or an LED.

The delivery can still differ because lasers and LEDs may have different:

• Beam profiles
• Divergence
• Treatment areas
• Coherence and spectral bandwidth
• Contact methods
• Pulse structures
• Power distribution across the treatment plane

Research standards apply to both lasers and LEDs. A study on lasers doesn’t automatically prove an LED product works, and vice versa. The device and protocol have to be close enough for the evidence to count.