Low-dose atropine for children: what it does, what it doesn't, and what to ask

Low-dose atropine drops do slow the progression of short-sightedness in children. The effect is real, it is modest, and it is still argued over — because two good trials have pointed in opposite directions about the weakest strength, and because regulators on either side of the Atlantic have looked at the same evidence and given different answers. The drops do not correct anything, so the glasses stay. They do not reverse anything either: an eye that has grown longer stays longer. What they can do is take some of the slope off the next few years, and whether they are doing that in your child is something you find out at the six-month review, not from the leaflet.

I should say at the start what I am and am not. I am a father who spent four years reading the research after my son’s eyes worried me; I am not a clinician, and both my children still see well. What follows is the account I would have wanted on the afternoon somebody read a minus number off a screen and handed me a card. The decision about whether and how to use these drops belongs to you and the person examining your child’s eyes. This page is for walking into that conversation knowing what to ask.

What the drops are, and what the trials actually tested

Atropine is an old drug. At full strength it is used to dilate the pupil for an examination. For myopia it is diluted enormously — the trials have used strengths between roughly a hundredth and a twentieth of one per cent — and its effect on eye growth runs through cell signalling that has nothing to do with the focusing muscle at all. In the trials, the routine was a single drop in each eye at bedtime. That is the whole of it, which is part of the appeal.

The concentration question is the whole argument. In the trial that mapped it most carefully — the one every later trial is compared against — children were given one of three dilutions or a placebo. The results lined up in a ladder: over the first year the children on the strongest of the three shifted by about a quarter of a dioptre, while the placebo children shifted by about four fifths of a dioptre. Stronger worked better. Stronger also brought more of the expected side effects, light sensitivity and a little blur close up.1 Evidence: STRONG

Then an American trial gave the weakest strength, 0.01%, to short-sighted children for two years and found it no better than placebo. Not slightly better: two hundredths of a dioptre apart, which is nothing at all.2 Evidence: STRONG

Two good trials pointing in opposite directions, so I am not going to pretend this is settled. The honest summary is that the weakest dilution has the shakiest record, the middle and stronger ones have done better in most places they have been tested, and nobody has a complete account of why the same drop behaves differently in different populations.1,2 Evidence: MODERATE

Why the regulators disagree, and what that tells you

The regulators have landed in different places too, which tells you something about the evidence.

In June 2025 a 0.01% drop was authorised across the whole European Union for children aged three to fourteen whose short-sightedness is progressing by at least half a dioptre a year. Its own pivotal trial, in more than eight hundred children, found a benefit of about an eighth of a dioptre over two years, and light sensitivity was the most common side effect.3 Evidence: STRONG That is a real benefit and a small one, and it is worth holding both halves of that sentence in your head at once.

In the United States there is still no approved low-dose atropine product at all. An application was turned back in October 2025 and sent for further review.4 Same molecule, same literature, two answers. If a clinic tells you this is simple, they have read less of it than you now have.

So: a modest, real, argued-over effect, from something a parent can actually do at bedtime. What it costs is about ten seconds a night for years, a child who will object to the drops for the first fortnight, and the mild nuisance of light sensitivity. In the book’s account it suits a young child, a child who will not tolerate anything in or on the eye during the day, and a family who want the treatment to be invisible at school. Which strength, and whether at all, is for the clinician — and the right question is not “should we use atropine” but “why that strength, for this child”.

When to start, and why six months is when you find out

Here is the part nobody explains in the appointment.

The thinking on timing has hardened. The International Myopia Institute now says that withholding treatment for long periods may no longer be considered ethical — unusually direct language for a professional body.5 Evidence: MODERATE The reason is arithmetic. The slope runs for years, the early years are the fast ones, and a year spent waiting to see what happens is a year of growth you cannot get back.

You will not know whether it is working in six weeks, or in three months. The review interval for a short-sighted child is six months, and that is the first point at which a change is big enough to mean anything. A year gives you a slope you can trust. The review should contain a refraction and a measurement of axial length — the length of the eye — plotted against the last one. Ask for the length, because it is the number the treatment is actually aimed at. About a tenth of a millimetre a year is ordinary growth; two or three tenths a year says the eye is progressing, whatever the letter chart says that morning.6 Evidence: STRONG

If the eye is still moving at six months, that is not a failure of you or of your child. It means one of three conversations: is the drop actually going in every night; is there a stronger version of what you are already doing; or is it time to combine two treatments. Combining is now a normal step rather than an exotic one. Adding low-dose atropine on top of an optical treatment has beaten either alone in several trials, and the Cochrane review found a further 0.12 millimetres a year of slowed growth from adding drops to night lenses.7 Evidence: MODERATE The optical options are compared in ortho-k vs. myopia-control lenses vs. glasses.

What happens when a child stops

This is the thinnest part of the evidence, and the Cochrane reviewers said so plainly: we have very limited evidence on what happens when these treatments stop.8 Evidence: LIMITED

For atropine specifically, what we have is sobering in one direction and encouraging in another. A five-year follow-up of the children in that first concentration trial found that, of those who stopped, nearly nine in ten had to restart because their eyes started moving again.9 Evidence: MODERATE Treatment ends when the eye stops growing, not on a birthday — and for most children that means into the mid-to-late teens, with the pace easing as they get older.

The encouraging part is about how you stop. The longest follow-up we have — eight years of the same children — compared stopping outright with tapering the drops down over three years. The tapered children ended up about a quarter of a dioptre better off, and nearly two thirds of them were still classed as good responders against fewer than half of those who stopped abruptly.10 Evidence: MODERATE Come off it gradually, under supervision, and be prepared to go back on. Ask about stopping at the start: the answer tells you how carefully the person in front of you has read the same literature.

The floor that every trial was built on

One thing before any drop, and it is not a treatment at all: the glasses must be right and must be worn.

A properly corrected lens does nothing to the length of the eye. What it does is put a sharp picture on the retina for every waking hour, which is the baseline every atropine trial was measured against — and which a startling number of children never get. When researchers pooled the studies on whether children wear the glasses they are given, the figure came out somewhere between four and five children in ten.12 Evidence: MODERATE Drops on top of glasses that live in a case on the windowsill are drops on top of nothing.

And refuse one thing if anybody offers it: a deliberately weak prescription so the eyes “have to work”. Children under-corrected by three quarters of a dioptre progressed faster, and the international clinical guidance now comes down to a single sentence — full correction, worn full time.11 Evidence: STRONG What the numbers on the card mean is in what the number on the prescription means.

Six questions to ask, in your own words

You will not remember these in the room, so keep them on your phone. They are the book’s six, and a good practitioner will not mind a single one.

1. “Are you measuring the length of my child’s eye, and can I see it on the chart?” If not, ask where locally it is done.
2. “How fast is this moving — in dioptres a year and in millimetres a year?” Speed decides everything else.
3. “Which treatments do you offer here, which would you choose for my child specifically, and why that one?” The why is what you are listening to — and if it is atropine, why that strength.
4. “What are we expecting to see at the six-month check, and what would count as this not working?” Agree the threshold before you need it.
5. “What do you do if it is still progressing — a stronger option, or do we combine?” You want to know there is a next move.
6. “What happens when we stop, and how do we stop?” Ask it at the start.

Then, on the evening of the appointment: put the card with the number on the fridge, because in six months that number is evidence. Book the review before you forget. Keep the two hours outdoors and the thirty centimetres going anyway — they are not the treatment, but they are free and still working on everyone else in the house. And tell your child, in the kitchen, that their eyes are growing slightly too fast, that there is something that can be done about the speed, and that you are going to do it. Children take these things at the temperature you serve them.

The alternatives to drops, and how they compare, are in the rest of the myopia-control hub — including the one treatment with the most striking numbers that I still cannot put on the list: red-light therapy.

Drawn from Chapter 14 of Children's Eyesight Problems, second edition. Report a correction.