My sister tells me that at night when the lights in the room are off that she can see the ambient outdoor light from neighboring buildigns and streetlights through her eyelids (i.e., when her eyes are closed). I, on the other hand, can't see anything. I can easily fall asleep even when the lights in the room are on; my sister can too, but when I tell her to sleep with the lights on because I have work to do, she just stares at the light and says that she can't sleep.

So, quantitatively, how much light would be able to pass through a normal human being's eyelids. And would ambient outside light be enough to pass through?

  • 3
    $\begingroup$ Define what you mean by "a normal human being". You seem to assume that you are, and your sister is not. But try an experiment: sit in a normally lighted room, close your eyes, and notice what you see. Now place your cupped hands over your eyes. Notice a difference? Then you're probably normal :-) (Though this might vary depending on the melanin content of your skin...) $\endgroup$
    – jamesqf
    May 12, 2021 at 16:49
  • 1
    $\begingroup$ Adding to @jamesqf 's personal experiment idea, note that visual sensitivity in humans spans many orders of magnitude an the dark-adapted eyes will notice smaller changes in light. Turn off the lights at night and sit by the switch for 5-10 minutes, then, keeping your eyes closed, flick the switch back on. I'd be surprised if you don't notice a difference, though what level of light is bothersome is surely going to vary from person to person. $\endgroup$
    – Bryan Krause
    May 12, 2021 at 18:27
  • 2
    $\begingroup$ Yep, this is a how-long-is-a-piece-of-string question. You could make some assumptions based on data on thickness of eyelids in the population and melanin content in general for specific populations- though that's a huge variable in itself. $\endgroup$
    – bob1
    May 12, 2021 at 20:57
  • $\begingroup$ @jamesqf by normal person, I mean just any average joe. Like most experiment would have an average reading right? $\endgroup$
    – prata
    May 13, 2021 at 4:43
  • $\begingroup$ @Brien Lim: Yes, but without doing experiments (or being much better at using search engines than I am) on a bunch of average Joes (or Jills, since your sister seems to have markedly different sensitivity), how can this be answered? And if we find that in fact there's a lot of difference between individuals, who's to say what's "normal"? $\endgroup$
    – jamesqf
    May 13, 2021 at 16:25

1 Answer 1


Short answer

About 0.79% of the light gets through on average, but there is variability across wavelengths and individuals. (variability between individuals seems to be correlated with density of macromolecules in the eyelid).

Long answer:

This is called "eyelid spectral transmittance".

Lucky for you, Bierman et al (2011) sought to determine the answer to this question.

Using a custom LED wand that actually emitted light under the eyelid for a sensor on the outside to read, they found quite a bit of variability both across wavelengths as well as across 27 test subjects.

They report their findings in terms of optical density, which is explained as follows:

A measure of the transmittance through an optical medium. Optical density equals the log to the base 10 of the reciprocal of the transmittance.... In spectroscopy, optical density is the measure of absorbance, and is defined as the ratio of the intensity of light falling upon a material and the intensity transmitted. Abbreviation OD.

We can use the equation A = -log10(1/T) to convert their findings of OD to transmittance and then multiply by 100 to get a percentage of light that can pass through.

  • Their result of an OD of 2.1 ± 0.3 SD = a transmittance of 0.79%

They do note that eyelid thickness, and/or the density of macromolecules in the eyelid account for most of the variation in transmittance between subjects.

Bierman, A., Figueiro, M.G. and Rea, M.S., 2011. Measuring and predicting eyelid spectral transmittance. Journal of biomedical optics, 16(6), p.067011.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .