Why 670nm? The Wavelength Behind Retinal Light Therapy

670nm is the deep-red wavelength most closely associated with photobiomodulation for the retina, because it sits at a sweet spot where the light is absorbed well by the energy machinery of retinal cells and still reaches the back of the eye. When researchers and device makers talk about red light for vision and retinal aging, 670nm is usually the number behind the claim. This page explains why that specific wavelength matters, what the research actually shows, and where the hype outruns the evidence.

It is a science explainer, not a device pitch. For the devices that use this wavelength and how they score, see our eye-device buying guide. We sell no hardware.

Why 670nm specifically?

670nm matters because of where it lands in the body’s response to light. It is a deep-red wavelength inside the optical window, the 600 to 1000nm band where light penetrates tissue relatively well and is absorbed by mitochondrial targets rather than scattered or blocked. Within that window, 670nm is absorbed efficiently by cytochrome c oxidase, the mitochondrial enzyme thought to drive photobiomodulation (de Freitas & Hamblin, 2016).

The retina is an unusually good target for this. It is one of the most energy-hungry tissues in the body, dense with mitochondria, and that demand rises as the eye ages. A wavelength that can nudge mitochondrial energy production is therefore of particular interest for the retina, which is why 670nm became the focus of retinal PBM research rather than a wavelength chosen at random.

The research behind 670nm

The most cited human work comes from University College London, where researchers led by Glen Jeffery tested brief 670nm exposure on aging eyes. In one study, a single short exposure to 670nm light improved colour-contrast sensitivity by up to 20% in people over 40, with no effect in younger eyes whose mitochondria were not yet declining (Shinhmar et al., 2021).

The age-dependence is the interesting part, because it fits the mechanism: if 670nm works by supporting tired mitochondria, it should help older retinas more than young ones, and that is what the data showed. Follow-up work from the same group explored timing, suggesting morning exposure mattered, which again points to a real biological effect rather than a placebo. These are small studies, but they are coherent and they used the 670nm wavelength specifically.

How 670nm reaches the retina

For a wavelength to help the retina, it has to actually get there, through the cornea, lens, and the fluid-filled interior of the eye. Deep-red light around 670nm passes through these structures better than shorter, bluer wavelengths, which are more strongly absorbed and scattered on the way in. That transmission is part of why 670nm, rather than a shorter red, is favoured for retinal work.

It is also why 670nm is considered gentle in this context: red and near-infrared light does not carry the photochemical-damage risk that intense blue or ultraviolet light does, and a review of ocular photobiomodulation reported no adverse effects at typical therapeutic doses (Gaspari et al., 2024).

670nm versus other wavelengths

670nm is not the only useful wavelength, and many devices pair it with others. Near-infrared light around 810 to 850nm penetrates deeper and is studied for tissue beyond the retinal surface, while clinic systems sometimes combine several wavelengths to hit more than one target. The reason 670nm holds a special place for the retina is the combination of good mitochondrial absorption and good ocular transmission, plus the body of human research built specifically around it.

The key point is that wavelength is necessary but not sufficient. A device emitting 670nm is using the right colour of light, but the dose still has to be right, because photobiomodulation follows a biphasic curve where too much cancels the benefit (Huang et al., 2009). A correct wavelength at the wrong dose is not a shortcut to results.

What 670nm does not do

It is worth being clear about the limits. The 670nm contrast-sensitivity research measured a specific visual function in a controlled setting; it is not evidence that the wavelength reverses eye disease or restores lost vision. The most robust clinical evidence for a retinal condition belongs to multi-wavelength clinic systems studied as whole devices, not to 670nm alone in a home gadget.

So the honest framing is that 670nm is a well-chosen, well-researched wavelength with a plausible mechanism and encouraging early human data for retinal aging. It is a promising tool, not a proven treatment, and any device leaning on the UCL research should be read as borrowing the wavelength’s evidence, not proving its own.

Frequently asked questions

Why is 670nm used for the eyes?

Because it is absorbed well by the mitochondria that power retinal cells and still penetrates to the back of the eye, while avoiding the damage risk of shorter wavelengths. That combination, plus human research built around it, made 670nm the focus of retinal photobiomodulation (de Freitas & Hamblin, 2016).

What does the 670nm research actually show?

A single short 670nm exposure improved colour-contrast sensitivity by up to 20% in people over 40, with no effect in younger eyes (Shinhmar et al., 2021). It is encouraging early evidence for retinal aging, not proof of disease reversal.

Is 670nm better than 810nm?

They do different jobs. 670nm red is favoured for the retinal surface, while 810nm near-infrared penetrates deeper, and some devices combine them. Neither is universally better; the right choice depends on the target and the dose.

Does a 670nm device guarantee results?

No. The correct wavelength still needs the correct dose, because the effect is biphasic and too much light cancels the benefit (Huang et al., 2009). Wavelength alone is not a shortcut.

Is 670nm light safe for the eyes?

Used as intended, it has a good safety record, with a review reporting no adverse effects at typical therapeutic doses (Gaspari et al., 2024). See our eye-safety guide for contraindications.

The bottom line

670nm earned its place in retinal light therapy by combining strong mitochondrial absorption, good penetration to the back of the eye, and a coherent body of human research showing improved contrast sensitivity in older eyes. It is the wavelength to look for in a retinal-focused device, but it is not magic: the dose still has to be right, and the evidence supports it as promising rather than proven. To see which devices deliver 670nm and how they compare, read our eye-device buying guide, and for the broader science, our photobiomodulation overview.

This guide is educational and not medical advice. It does not create a doctor-patient relationship. Always talk to a qualified eye-care professional before starting any light-based therapy.