Look at the two orange circles in the figure below. Same size? They are. You can check by dragging the slider, and then check again by switching the rings off. Your eyes still won't quite believe it.
The circle on the left sits inside a ring of big circles. The circle on the right sits inside a ring of small ones. The left circle is fixed. Resize the right one until the two look equal to you, then lock in your answer.
If you made the right circle noticeably smaller than the left one, the illusion worked on you. If you landed close to 100%, the illusion was weak for you, which happens (see below).
This figure is called the Ebbinghaus illusion, after the German psychologist Hermann Ebbinghaus. Edward Titchener put it in an English-language psychology textbook in 1901, which is why you will also see it called the Titchener circles.
The effect is simple: a circle among big circles looks smaller than the identical circle among small ones.
Your visual system does not read off sizes like a ruler. It judges size relative to what is nearby. A circle next to giants gets read as small; the same circle next to tiny dots gets read as big. The basic explanation is called size contrast.
That usually serves you well. In a real scene, comparing an object with its neighbours is a good way to guess its size when you do not know how far away it is. The illusion is that same shortcut firing in a drawing where it gives the wrong answer.
Move the gap slider above and watch what happens. Researchers have found that the distance of the surrounding circles matters too: ring circles that sit close to the centre tend to make it look larger, and ring circles that sit farther away tend to make it look smaller. Big surrounding circles have to sit farther out, so in many classic drawings size and distance are mixed together. Whether the ring is a complete circle also matters, which links this figure to a cousin called the Delboeuf illusion.
So "big neighbours shrink you" is only part of the story. The layout of the whole scene counts.
People differ in how strongly the illusion works. A 2011 brain-imaging study by Schwarzkopf, Song and Rees found that the more susceptible a person was, the smaller their primary visual cortex (V1) tended to be: an inverse correlation between susceptibility and the surface area of that early visual region.
Age matters as well. In a 2010 study by Doherty and colleagues, adult students were fooled more than children aged ten and under. One reading of that result is that taking in the whole scene, context and all, is a skill that develops, and that younger children look more at each shape on its own.
Does the illusion fool your hand as well as your eyes? Researchers disagree. One study using a physical version of the figure found that the way people open their fingers to grasp the centre disc was barely affected, which was taken as a clue that seeing and acting use different routes in the brain. Other researchers reported that grasping is fooled too. It is still an open debate.
Next time two things look different in size, ask what is around them. Designers use this on purpose: a button looks larger among small text, a room looks bigger beside a tiny one. And when you need the truth, measure. Your eyes are brilliant at telling you how things compare, and not at all built to be a ruler.