Where the colour wheel came from, and why there are really two

The colour wheel wasn't invented once, and there isn't just one of it. Isaac Newton published a wheel of light in 1704, not 1666 as plenty of pages claim. Whereas painters started their own red, yellow and blue wheel in 1708, and Johannes Itten turned that one into the school version in 1961. The two wheels still disagree about which colours sit opposite each other. That's why your design tools can give you different "complementary" colours for the same starting colour.

Who invented the colour wheel?

Nobody did, at least not in one go. A 1982 paper with the blunt title "Who invented the color wheel?" (opens in a new tab) traces colour circles back to at least 1611, and they come in several different families. The red, yellow and blue circle you met in art class is actually the youngest of them, from the 1700s. For about 2,000 years before that, people usually drew colours in a line from white to black, an idea that goes all the way back to Aristotle.

There were colour circles before Newton, but none of them was a circle of the rainbow's hues. Newton's came first, in his book Opticks, in 1704. Four years later the painters started one of their own. And those two wheels, one for light and one for paint, have never fully agreed.

Worth knowing: the "1666" date is wrong. Plenty of pages say Newton invented the colour wheel in 1666, but that's the year he says he bought a prism, in a letter published in 1672 (opens in a new tab). His notebook, his lectures from about 1670 and that letter don't contain a circle at all. The circle was first published in Opticks (opens in a new tab), in 1704.

Newton's wheel was a calculator for light

It was never meant to be a palette. Newton took the band of colours a prism throws on a wall, red through to violet, and bent it into a ring of seven slices with white in the middle. Then he used it to answer one question: if you shine these lights together, what colour do you get?

The way it works is a balancing act. Picture the wheel as a see-saw with weights round the rim, and put a weight on each colour, heavier for more of that light. The spot where the wheel would balance tells you the mix. The direction from the centre gives you the colour, and the distance gives you its strength. Near the rim it's strong, near the middle it's pale, and dead centre it's white. Newton himself admitted the rule was close enough for practice but "not mathematically accurate".

Now, bending the rainbow into a ring left a gap where red meets violet. Newton noticed that mixing light from those two ends gives a purple, which is a colour the prism never shows. That's why the purples and magentas on every colour wheel are mixtures, not rainbow colours.

He also knew paint doesn't follow his rule. When painters mixed their powders, he wrote, they got dull greys and browns, the colour of mouse fur, ashes or road dust.

A side note on indigo: it's in the rainbow because of music. In his earlier work Newton used five main colours: red, yellow, green, blue and purple. Later he added orange and indigo to make seven, so they'd match the notes of a musical scale. He even sized his slices to musical steps rather than to how the colours look, which is why red, green and violet got the biggest slices and orange and indigo the smallest.

The painters' wheel started in 1708

It turns up in a French manual on miniature painting (opens in a new tab) printed in The Hague. The book is signed "C.B." and often credited to Claude Boutet, and it holds the earliest known painters' colour circles: one in seven parts and one in twelve, both coloured by hand. Both are built on yellow, red and blue, the three paints the book says every other colour comes from.

But then the book admits one red isn't enough. Painters need two of them: a fire-red that leans towards yellow, and a crimson that leans towards blue. So if you keep a warm red and a cool red on your palette today, you're following advice from 1708. What the manual didn't do was make anything special of opposite colours. Mix them together, it said, and you just get dirty colours.

Opposites only became the point of the wheel about 60 years later. Around 1770 Moses Harris published The Natural System of Colours, with wheels built on red, yellow and blue, each colour fading from strong at the rim to pale near the centre. He set the opposites across the circle: red and green, blue and orange, yellow and purple. And his proof came from the laundry. Washerwomen added blue to whiten linen that had gone yellow. He'd also noticed that after you look through green spectacles, the world looks reddish. Harris's wheel is the ancestor of the red, yellow and blue wheel still taught in schools.

You'll often see 1766 given for Harris, including by the US Library of Congress. The Royal Academy (opens in a new tab) calls that date mistaken, and the clue is in the dedication. The book is dedicated to Sir Joshua Reynolds, who wasn't knighted until May 1769, and one of its circles turned up again in 1776. So about 1770 is the safe date.

Why do paint and light mix so differently?

Because paint takes light away, whereas light adds it up. Every paint acts as a filter: it blocks some of the light, so each paint you add makes the mix darker. Light goes the other way. Shine two torches at the same spot and it gets brighter, and with enough of them you get white.

People spotted the split early on. In 1725 the printmaker Le Blon wrote that yellow, red and blue paints mixed together make black, while all the colours of light mixed together make white, as Newton had shown.

The proof had to wait until 1852, when the German scientist Helmholtz showed that yellow light and blue light make a faint greenish white, not green. He pointed out that this went against a thousand years of painters' experience. Paint behaves differently because blue paint lets through blues and greens, and yellow paint lets through yellows, reds and greens. So when you mix the two, green is the only light that gets past both.

The painters' wheel still works for paint for a related reason. The "red" and "blue" painters mix with are rough stand-ins for magenta and cyan, which are the mixing colours printers use. Why bright designs go dull in print picks up that side of the story.

Itten's Bauhaus star isn't the school wheel

No, it isn't. Johannes Itten taught at the Bauhaus art school from 1919 to 1923, and his famous 1921 colour star has 12 hues. Those hues include a cyan-blue and a purple-red, though, with yellow at the bottom. On that chart vermilion sits opposite cyan-blue, and yellow sits opposite blue-violet. Those pairs are closer to the light wheel's opposites than to the painters'.

Itten got the 12-step circle from his teacher, Adolf Hölzel, who was already using it by 1904. And Hölzel's order came from the physicist Bezold, back in 1874. The tidy school wheel came much later, in Itten's 1961 book The Art of Color: yellow, red and blue, then orange, green and violet, then six in-betweens, all evenly spaced. That's the one on the classroom wall.

Why do design tools give different complementary colours?

Say you pick red in Canva and ask for its complement, and you get cyan. Then you try Adobe Color, and by an Adobe employee's account you'd expect green. The two tools are answering on different wheels.

The wheel in most digital colour pickers comes from Alvy Ray Smith's 1978 hue, saturation and value model. That's a wheel of light, where hue is an angle: red at 0°, yellow at 60°, green at 120° and blue at 240°. Painters and some design tools still use the red, yellow and blue wheel, though. Canva says its colour wheel is an RGB wheel, built for screens. An Adobe employee has said Adobe Color's harmony rules run on the red, yellow and blue wheel. And Paletton says it uses the "artistic RYB" wheel. Ask each one for red's opposite and you'll get different answers.

Red and yellow are where the wheels disagree most. In my own test, using one published recipe for drawing a painters' wheel on screen, a mid blue got almost the same hue of orange on both wheels, although the screen wheel's orange is darker and browner. The same recipe gives the stretch from red to yellow a third of the painters' wheel, but only a sixth of the screen wheel. In other words, warm colours get twice the room on the painters' wheel.

That doesn't make either wheel wrong. They were drawn for different jobs, and the mistake is only using the wrong one for yours. Whether opposite colours actually look good together is a separate question, and that's covered in Do colour harmony rules actually work?

What should you do differently in your next design?

Pick the wheel that matches the job, and don't treat either one as a rule for taste.

1. Check which wheel a tool uses before you trust its "complementary" colour. Look for "RGB" or "RYB" in its help pages. If two tools disagree, that doesn't mean either one is broken.
2. For screen work, think in light. A screen mixes red, green and blue light, so a screen wheel describes what your design will actually do.
3. For paint, use the painters' wheel. It has guided paint mixing since at least 1708. And if you mix your own, keep a warm red and a cool red.
4. Treat "opposite" as a starting point, not a verdict. Neither wheel tells you what looks good, so judge the pair in your actual design.
5. Retire "Newton, 1666" from your workshop slides. Newton's wheel was first published in 1704, and it was a wheel of light.

Frequently asked questions

Who invented the colour wheel?

No single person did. Isaac Newton published the first circle of rainbow colours in his book Opticks in 1704, not 1666, which is just when he bought his prism, and his wheel described how light mixes. The painters' red, yellow and blue wheel started separately in a French painting manual of 1708, and Moses Harris built on it around 1770, then Johannes Itten in 1961. Simpler colour circles go back to at least 1611.

What's the difference between an RYB and an RGB colour wheel?

They describe two different kinds of mixing. An RYB wheel is the painters' wheel, built on red, yellow and blue paint, where mixing makes colours darker. An RGB wheel describes light, as on a screen, where red, green and blue add up to white. Because paint and light mix differently, the two wheels put different colours opposite each other: red faces green on an RYB wheel but cyan on an RGB wheel.

Why do Canva and Adobe give me different complementary colours?

Because they use different colour wheels. Canva says its wheel is an RGB wheel, built for screens, so the opposite of red is cyan. An Adobe employee has said Adobe Color's harmony rules run on the traditional painters' wheel of red, yellow and blue, where the opposite of red is green. Neither one is broken. They're answering the same question on two different wheels, so check which one a tool uses before you trust its answer.

Is the red, yellow and blue colour wheel wrong?

Not wrong, but it is limited. It's a practical guide for mixing paint, and painters have used it since at least 1708. Where it goes wrong is as a law of colour: no three paints mix every colour, printers get more from cyan, magenta and yellow, and screens use red, green and blue light. Albert Munsell was calling red, yellow and blue as the primaries "a widely accepted error" as early as 1907.

Why is purple on the colour wheel but not in the rainbow?

Because the rainbow is a line, not a circle. It runs from red to violet and then stops. When Isaac Newton bent it into a circle in 1704, he joined the two ends, and he noticed that mixing light from them gives a purple the prism never shows. So the purples and magentas on every colour wheel are mixtures of the rainbow's two ends, not single rainbow colours.

For the curious

The two kinds of mixing are additive (light, which adds up towards white) and subtractive (paint and filters, which take light away). The screen wheel is the RGB or HSV wheel (Smith, 1978). The painters' wheel is RYB. Printers mix with CMY: cyan, magenta and yellow.

The web's colour standard, CSS Color 4 (opens in a new tab), says these hue angles are not evenly spaced to the eye. Why "same brightness" colours don't look the same explains why.

Newton's balance-point rule is his centre of gravity method, in Opticks, Book I, Part II, Proposition VI. In 1704 the seven slices are 1/9, 1/16, 1/10, 1/9, 1/10, 1/16 and 1/9 of the circle, and those musical steps make exactly one octave. In later editions, from 1718 on, blue is printed as 1/16, which breaks the octave.

Software can't mix real paint, so any on-screen RYB wheel is an approximation. The recipe behind my blue and warm-colour comparisons is Gossett and Chen's (2004). It picks eight corner colours by eye, based on Itten, and blends between them. Its authors say it isn't based on any model of perception, and other tools may well use other recipes. If you want to go further, David Briggs's "The RYB hue circle" (opens in a new tab) is an excellent free deep dive.

Sources (23)

• Isaac Newton, Opticks (opens in a new tab), 1st edn, London, 1704, Book I, Part II, Prop. V and Prop. VI, pp. 114–117
• Isaac Newton, "A letter … containing his new theory about light and colors" (opens in a new tab), Philosophical Transactions 6, 1671/2 (The Newton Project)
• A. E. Shapiro, "Experiment and mathematics in Newton's theory of color" (opens in a new tab), Physics Today 37(9), 1984
• C. Parkhurst and R. L. Feller, "Who invented the color wheel?" (opens in a new tab), Color Research & Application 7(3), 1982 (abstract)
• Aristotle, On Sense and the Sensible (opens in a new tab), tr. J. I. Beare
• [C.B., often credited to Claude Boutet], Traité de la peinture en mignature (opens in a new tab), The Hague, 1708
• J. C. Le Blon, Coloritto, or the Harmony of Colouring in Painting (opens in a new tab), London, [1725], p. 6
• Moses Harris, The Natural System of Colours, c. 1770, transcribed by J. L. Caivano (opens in a new tab)
• Royal Academy of Arts, The Natural System of Colours (opens in a new tab), collection record
• H. Helmholtz, "Ueber die Theorie der zusammengesetzten Farben" (opens in a new tab), Annalen der Physik, 1852
• A. H. Munsell, A Color Notation (opens in a new tab), 2nd edn, 1907, appendix "False Color Balance"
• J. Itten, "Farbenkugel in 7 Lichtstufen und 12 Tönen", plate in Utopia: Dokumente der Wirklichkeit (opens in a new tab), Weimar, 1921
• J. Itten, The Art of Color (opens in a new tab), 1961
• U. Röthke, "Hölzels Farbenlehre im Kontext seines Kunstunterrichts" (opens in a new tab), Kunstgeschichte Open Peer Reviewed Journal, 2011
• Bauhaus Kooperation, Johannes Itten's preliminary course (opens in a new tab)
• A. R. Smith, "Color gamut transform pairs" (opens in a new tab), SIGGRAPH '78, 1978
• W3C, CSS Color Module Level 4 (opens in a new tab), Candidate Recommendation Draft, 13 September 2026
• N. Gossett and B. Chen, "Paint inspired color mixing and compositing for visualization" (opens in a new tab), IEEE InfoVis, 2004
• Encyclopaedia Britannica, "Subtractive mixture" (opens in a new tab)
• D. Briggs, The Dimensions of Colour, "The RYB hue circle" (opens in a new tab)
• Canva, colour wheel page (states its wheel is RGB), retrieved 19 September 2026
• Adobe Community, "How is complementary color determined?", answer by an Adobe employee, 9 March 2023
• Paletton, home page (states it uses an RYB wheel), retrieved 19 September 2026