The 20-20-20 Rule: What It's For, and What It Can't Do

Every twenty minutes, look at something twenty feet away, for twenty seconds. It is the most repeated piece of eye advice on the internet, it is genuinely sensible, and almost every article that recommends it gets one thing wrong: it does not say what the rule is for.

That omission matters, because people adopt 20-20-20 for symptoms it does not treat, conclude it does not work, and drop it.

Where the numbers come from

The rule is widely attributed to the American optometrist Jeffrey Anshel, who proposed it as something patients would actually remember. The memorability is the design goal — it is a mnemonic first and a dose second.

One of the three numbers is not arbitrary, though. Twenty feet — about six metres — is roughly where optical infinity begins for the human eye. Beyond that distance, incoming light is close enough to parallel that your ciliary muscle can relax essentially completely. That is the whole point of the number: it is the distance at which your focusing system stops working.

The other two numbers are plausible estimates rather than findings. No trial established that twenty minutes is the correct interval, and the one trial that has tested the twenty-second break found it too short to do anything.

What it treats

Sustained near work means holding a ciliary muscle contraction for hours. Like any static muscular effort, it accumulates fatigue and it stops releasing cleanly. The visible consequence is distance blur after long reading sessions — you look up and the far wall takes a moment to arrive.

For that, 20-20-20 has a clear mechanical rationale. A genuine distance fixation lets the muscle release before the effect builds. It is the right intervention, aimed at the right mechanism.

It also does something the mechanism does not capture: it makes you look up. Standing, changing posture and shifting attention address the neck and the tension headache, which are a substantial share of what people call eye strain in the first place.

What the evidence actually says

Less than the ubiquity suggests — and the one study that set out to check the numbers directly did not support them.

Johnson and Rosenfield opened by noting that there is relatively little peer-reviewed evidence for the rule at all, then tested it: thirty subjects, a 40-minute tablet reading task, 20-second breaks allowed every 5, 10, 20 or 40 minutes. The breaks produced no significant effect on reported symptoms, on reading speed, or on accuracy. Their conclusion was that a 20-second break does not work as a therapeutic intervention for digital eye strain, and that a useful one is probably closer to one or two minutes.

Johnson, S. & Rosenfield, M. (2023). 20-20-20 Rule: Are These Numbers Justified? Optometry and Vision Science, 100(1), 52–56.

That result deserves to be stated plainly: the twenty-second part of the rule has been measured, and it did not help. The mechanism behind it is still sound — releasing a sustained accommodative contraction is a real thing to want — and the intervention is free and risk-free, which is why it is still worth doing. But "recommended everywhere" and "well evidenced" are different claims, and 20-20-20 has only the first. If you are going to stop, stop for longer than twenty seconds.

The gap nobody mentions

Here is the arithmetic that undermines 20-20-20 as general eye care:

The two dominant complaints from screen work have different mechanisms. Focusing fatigue is a muscular problem, and a break fixes it. Dry eye is an evaporation problem driven by a blink rate that drops by roughly two thirds at a screen, along with blinks that never fully close.

A break does not repair a tear film for the period after the break. You return to the screen, the blink rate collapses again within seconds, and the film destabilises on the same timescale as before. This is why people with burning, gritty eyes often report that 20-20-20 "did not do anything" — accurately. They were treating a different organ.

The interventions are complementary, not interchangeable: 20-20-20 for the ciliary muscle, blink prompts for the ocular surface. Doing one and expecting both is the most common mistake in this whole subject.

The real failure mode is adherence

Ask a room of software engineers whether they know the 20-20-20 rule and every hand goes up. Ask how many followed it yesterday and the hands go down.

The reason is structural rather than motivational. Twenty minutes is precisely the timescale of a deep work block. The rule asks you to interrupt yourself at the moment interruption is most expensive, using the attentional system that concentration has switched off. Intention is not the constraint; the task is.

Two consequences follow, and both are practical:

• The prompt has to be external. Deciding to remember does not survive contact with a working afternoon. This is the same finding that underlies blink training: internal resolve fails within seconds of returning to the task.
• The prompt has to know when to shut up. A break timer that fires while you are presenting to forty people gets uninstalled that afternoon. Not because it is wrong, because it is unusable.

Making it work in practice

• Make the twenty feet real. Looking at a wall two metres away is not a break — the ciliary muscle is still engaged. Pick a window, a corridor, or the far end of the room, and use the same target every time so you do not have to decide.
• Stand up for at least some of them. It costs nothing extra and it addresses the neck, which is doing more of the damage than most people credit.
• Take longer than twenty seconds when you can. The evidence says a minute or two does what twenty seconds does not. Twenty seconds is the version you will actually do at your desk; treat it as the floor, not the target.
• Do not chase perfect compliance. Twelve breaks in an eight-hour day is a good day. Twenty-four is a fantasy that ends with the timer switched off.
• Pair it with a blink prompt. If your symptoms are burning, grittiness or vision that clears when you blink, breaks alone will not touch them.

How Skopia implements it

Skopia runs the rule as a background timer with a 20-second window, alongside a separate palming break every 60 minutes — two minutes with your palms cupped over closed eyes. Both live in the Rest tab and can be switched off independently.

Three implementation details are opinions, not defaults we inherited:

• Only one timer runs at a time. While you are inside a rest window, the other counters freeze. Stacked prompts are how break apps become noise.
• Rest windows reset the blink countdown. You have just spent twenty seconds looking into the distance and blinking normally; being told to blink four seconds later is absurd.
• Everything freezes when you should not be interrupted. On a call, recording your screen, in full screen, or in an app you have marked as deep focus — the counters hold and resume from where they stopped. A break you skipped is not a break you owe.

The rule is good. It is just narrower than its reputation, and it only helps on the days you actually follow it.