Lighting for eye strain: what the sources actually say

Search this topic and you will get a confident number within about four seconds. 500 lux. Or 300–500. Or 750. Usually attributed to a standards body, usually with no link, and usually stated as though the matter is settled.

It is not settled. We went and read the primary sources, and the two national bodies most often invoked — OSHA in the United States and CCOHS in Canada — publish recommendations for a computer desk that differ by roughly a factor of 2.6. Neither is wrong. They are answering the question from opposite ends.

That disagreement is the most useful thing on this page, so we are going to lead with it rather than smooth it over into a number we made up.

What OSHA actually published

First, a unit problem that matters more than it sounds like it should. OSHA does not publish lux. It publishes foot-candles. Verbatim, from its computer-workstations eTool:

“office lighting should range between 20 to 50 foot-candles (minimum of 30 foot-candles for employers in the construction industry). If LCD monitors are in use, higher levels of light are usually needed for the same viewing tasks (up to 73 foot-candles).”

You will see that rendered elsewhere as “215–538 lux (OSHA)” with no further comment. It is worth being precise about why that is a problem, because it is subtle and it is everywhere.

One foot-candle is 10.764 lux. So our conversionof OSHA’s range gives roughly 215–538 lux, and our conversion of the 73 fc LCD figure gives roughly 785 lux. Those are honest arithmetic. But the moment you print them as OSHA’s figures, you have attributed to a federal agency a set of numbers it never published, in a unit it does not use, with a false precision the original never claimed. “20 to 50” is obviously a rough band. “215–538” looks like it came out of an instrument.

So: the lux values on this page that trace to OSHA are ours, they are labelled as ours every time they appear, and the foot-candle figures are the ones OSHA stands behind.

What CCOHS actually published

CCOHS — the Canadian Centre for Occupational Health and Safety — publishes in lux, freely, which is why it is our source for the lux side of this page rather than the IES.

Its recommended-illumination table puts computer tasks at 75–300 lux. That is a low number, and if you have absorbed the usual internet wisdom it may be a startling one.

Its general office figure is the famous 300–500 lux — and CCOHS attaches two qualifications to it that essentially never survive being quoted. First, the recommended level “is not a must.” Second, “it applies in the situation where there is no task lamp in use.” Both of those are CCOHS’s own words about CCOHS’s own number, and dropping them turns a hedged guideline into a specification.

CCOHS also gives a general lighting figure of 500–1000 lux measured 76cm (30 inches) above the floor— note the stated measuring height, which is the kind of detail that makes a light figure meaningful and which most sources omit entirely.

One point of provenance, since this is where most pages quietly cheat. The 300–500 figure is usually credited to the Illuminating Engineering Society, generally to RP-1. We have not read RP-1 — it is paywalled— so we do not cite it. CCOHS names the IESNA Lighting Handbook (9th ed.), CSA Z412:24 and British Columbia’s OHS Regulations as the sources behind its own table. We can tell you CCOHS cites those. We cannot tell you what is in them, and we are not going to borrow their authority by implying we looked.

The conflict, stated plainly

Line the two up for the same desk — a person at a computer:

• CCOHS:computer tasks, 75–300 lux.
• OSHA:where LCD monitors are in use, up to 73 foot-candles — about 785 lux by our conversion.

The top of CCOHS’s range is 300. OSHA’s LCD ceiling is roughly 785. That is not a rounding difference or a translation artefact. Two national occupational safety bodies are describing the same desk and landing about 2.6x apart.

We could have hidden this. Pick whichever range flatters the products, print it as fact, move on — nobody checks. Instead: here are both, with links, and here is our best read of why they differ.

Why they differ: the monitor is a light source

The reason the question splits is that a screen is not paper.

Paper is passive. To read it, you have to put light on it, and more light makes it easier to read until glare sets in. That is the world the old office-lighting numbers were built for, and it is why a document-heavy desk wants a decent task lamp.

A monitor supplies its own light. You do not need to illuminate it — in fact, light landing on the panel actively degrades it, washing out the image with reflected glare. So from the screen’s point of view, the ideal ambient level is low. That is a coherent way to reach CCOHS’s 75–300 lux.

But from your eyes’ point of view, a bright self-lit rectangle in a dim room is its own problem, and this is where OSHA’s reasoning lives. Its LCD sentence says higher levels are needed “for the same viewing tasks” — the surrounding room has to come up to meet the panel. That is a coherent way to reach 73 foot-candles.

Two defensible positions, optimising different halves of one problem. Which is why the single confident number you have been given elsewhere is a choice someone made without telling you.

OSHA’s real mechanism isn’t brightness at all

Here is the part that reframes the whole topic, and it is the sentence we would keep if we could keep only one. From the same OSHA page:

“High contrast between light and dark areas of the computer screen, horizontal work surface, and surrounding areas can cause eye fatigue and headaches.”

The subject of that sentence is contrast. Not lux. Not “too dark” or “too bright” in absolute terms — the gap between the bright thing and the dark thing next to it.

This explains why the lux argument is less decisive than it looks. A desk at 200 lux with a dimmed monitor and a desk at 700 lux with a bright monitor can both be comfortable, because in each case the gap is small. A desk at 50 lux with a monitor at full brightness is a lit rectangle in a cave, and no lux target on its own would have told you that.

It also reframes the shopping question. If the mechanism is contrast, then the cheapest intervention is not a light at all — it is turning your monitor down. That closes the gap from the other side, costs nothing, and nobody selling lamps has any reason to mention it.

The free things, first

Before any of this becomes a purchase, the sources point at several things that cost nothing.

The 20-20-20 rule.CCOHS states it verbatim: “At least every 20 minutes, take a 20-second break and look at something 6 metres (20 feet) away.” It addresses focus distance, not light level. Your focusing muscles hold one position for hours while you look at a panel about 60cm from your face; the rule interrupts that. It is free.

Monitor brightness. If contrast is the mechanism, the panel is half of the equation and it has a slider.

Where the light already is.CCOHS’s guidance on positioning is straightforward — place the monitor parallel with overhead lights rather than directly beneath them, angle it away from lights and windows, and make sure any task lamp illuminates the document and not the monitor. Rotating a desk 90 degrees is free and fixes more glare than most products.

Where a product might legitimately help

If the gap is genuinely open — you have looked at your desk in the evening and the surface in front of the monitor is obviously darker than the screen — then raising the light around the panel without putting light on the panel is a real, narrow, geometric job.

That is what a monitor light bar is built to do: it clamps above the bezel and aims down and forward, so the desk comes up and the glass stays clean. For scale, BenQ’s ScreenBar Pro publishes 85 × 50cm of coverage at 500 lux — comfortably inside every range on this page.

If you also look at physical objects — paper, a notebook, your hands — then the passive-surface problem is back and you want a task lamp instead. Which is exactly the situation CCOHS’s hedge was pointing at: its 300–500 lux applies “where there is no task lamp in use.” Put a lamp on the desk and the ambient requirement changes. Most pages quote the number and drop the condition that governs it.

What this page is not

This is not medical advice and it is not written by anyone qualified to give any. Stephen is an enthusiast who reads standards documents and spec sheets. We have described what two occupational safety bodies published about light levels and contrast, quoted them directly, linked them, and flagged where they contradict each other and where our own arithmetic enters the picture.

We have not told you that any product will reduce your eye strain, because we do not know that and cannot know it. We have not tested anything. If your eyes hurt consistently, that is a conversation for an optometrist, who can check things about your vision that no lux figure addresses.

What we can offer is this: the numbers you have been quoted elsewhere are shakier than they look, the two biggest authorities disagree by 2.6x, the mechanism that actually gets named in the source material is contrast rather than brightness, and the first two fixes are free. That is a more honest starting point than a confident number, even if it is a less comfortable one.