Does astrology work? What happens when it is tested?

What do people observe?

Start with what people experience. Someone reads their horoscope, and it fits. An astrologer reads a birth chart, which is a map of where the Sun, Moon and planets were at the moment and place of a person’s birth, and says things that ring true. A bad week lines up with “Mercury retrograde.” These moments are real and common.

So is interest in astrology. In a Pew Research Center survey of 9,593 U.S. adults in October 2024, 28% said they consult astrology or a horoscope at least once a year, and 27% said they believe in astrology. That share has barely moved: it was 29% in 2017. Belief was about the same among Americans who have a religion (27%) and those who do not (28%).

Most people are not staking their lives on it. Pew found that 20% of U.S. adults use astrology, tarot or fortune tellers mostly “just for fun,” while 10% say these practices give them helpful insights. Only 1% said they rely a lot on them when making major life decisions.

What is inferred from it?

The inference is bigger than the experience. Astrology holds that the positions of the Sun, Moon and planets at your birth are linked to your personality and to events in your life. Sun-sign columns boil this down to 12 types, while astrologers read a whole chart. Both make the same basic claim: the sky at your birth says something true and specific about you.

That claim can be tested. If it is true, astrologers should be able to match charts to people better than chance, and people born at the same time and place should be more alike than strangers.

Does the evidence support it?

The Nature test. In 1985, the physicist Shawn Carlson, then at the University of California, Berkeley, published “A double-blind test of astrology” in Nature. For each of 116 birth charts, astrologers received three profiles from the California Psychological Inventory, a standard personality questionnaire. One profile was real; two were picked at random. The astrologers chose the right one first 40 times, or 34.5%. Guessing would give 33.3%. In a second part, 83 volunteers each read three chart readings, one based on their own chart. They did not pick their own more often than chance.

Time twins. The researchers Geoffrey Dean and Ivan Kelly, a professor of educational psychology at the University of Saskatchewan, called people born at nearly the same moment the “definitive test” of astrology. In a 2003 paper in the Journal of Consciousness Studies, they described data on 2,101 people born in London from March 3 to 9, 1958, on average 4.8 minutes apart. Measurements at ages 11, 16 and 23 covered 110 traits, from IQ scores and sociability to height and marital status. The astrologer John Addey had written that people born within a few minutes of each other in the same place should show “really exceptional” likenesses. The measured effect was zero.

Many smaller tests. The same paper pooled more than 40 studies in which nearly 700 astrologers tried to match about 1,150 birth charts to people. The combined result was not reliably better than chance. In ten studies with nearly 300 people, people could not pick out their own chart reading. And across 25 studies with nearly 500 astrologers, astrologers reading the same charts showed almost no agreement with each other.

Why horoscopes feel right. In 1949, the psychologist Bertram Forer gave 39 students a personality test. A week later, each got a typed sketch with their name on it. Every sketch was the same, and Forer wrote that its statements “came largely from a newsstand astrology book.” One line read: “You have a great need for other people to like and admire you.” On a scale of 0 to 5, the students rated how well it described them at 4.26 on average. Dean and Kelly list this effect, reading specific meaning into general statements, among the “hidden persuaders” that can make a reading feel like a hit. They also report that a link between sun signs and outgoing personality, found in 1978, vanished in later studies of people who knew nothing about astrology. Knowing your sign can nudge how you see yourself.

The sky has moved. NASA explains that when the Babylonians set up the 12 signs, a birthday from about July 23 to August 22 meant being born under the constellation Leo. Earth’s axis slowly wobbles like a spinning top, a motion called precession that takes about 25,770 years per cycle, according to NASA. After 3,000 years, the Sun no longer lines up with the same stars on the same dates. The Sun also passes through 13 constellations, not 12; the Babylonians left out Ophiuchus. NASA’s summary: “We didn’t change any zodiac signs, we just did the math.”

Mercury retrograde. Mercury never actually reverses. As NASA’s StarChild site explains, retrograde motion is an apparent change: a planet only seems to move backward because of how it and Earth move around the Sun. Mercury circles the Sun in 88 days, so it laps Earth three or four times a year and seems to go backward for about three weeks each time, the Planetary Society explains. It adds that there is no scientific evidence that any planet’s apparent retrograde motion affects you.

What is the strongest objection, and what do people get wrong?

“The famous test was unfair.” This is the most serious reply, and it deserves a hearing. In 2009, Suitbert Ertel of the University of Göttingen reanalyzed Carlson’s data in the Journal of Scientific Exploration. Carlson had said he planned to count first and second choices together. Counted that way, the astrologers got 86 of 116 right, against about 77 expected by chance. Ertel called this borderline, with a p-value of .054, a measure of how often chance alone would give a result this strong. He also quoted an astrologer who had advised the study, Teresa Hamilton, saying she was never satisfied it was a fair test. Ertel concluded that the results were not enough to count astrology as verified, but that Carlson’s strong negative verdict went too far. So even on the kindest reading, the best case is a small, disputed edge in one test, not the reliable skill astrology claims.

“What about the Mars effect?” The French statistician Michel Gauquelin reported that champion athletes and other eminent people were born a little more often than expected when some planets, such as Mars, were rising or at their highest point in the sky. Supporters such as Ertel say the pattern held up in new data. Dean and Kelly say it fits parents adjusting birth records to suit popular beliefs, which was easy to do undetected in the 1800s and early 1900s. Even Ertel treats it as separate from classical astrology. It concerns where planets sat relative to the horizon, not sun signs or horoscopes.

“Western astrologers use signs, not constellations.” This is fair. Most Western astrologers use the tropical zodiac, which starts Aries at the spring equinox and follows the seasons rather than the stars, according to the Encyclopedia of Occultism and Parapsychology. So the drift of the constellations does not change their charts. But the tests above did not depend on constellations. They tested charts as astrologers read them, and the charts did not predict.

A common mistake on the other side is to treat everyone who reads a horoscope as a true believer. Pew’s numbers show that most treat it as fun.

What is still unresolved?

No test can prove an effect is exactly zero. Dean and Kelly say plainly that researchers can never be sure about ruling out very tiny effects. What they can rule out is an effect as large as astrologers claim, the kind that would let a chart describe a person usefully. Ertel and Dean have disputed the Gauquelin data in print for years. And why so many people find astrology fun or helpful, which Pew’s survey only begins to measure, is a real question for psychology. It is separate from whether astrology predicts anything.

How can you check it yourself?

• Swap the horoscopes. Ask friends to read this week’s horoscopes with the sign names hidden, then pick their own. If astrology works, they should do better than one in 12.
• Try Forer’s test. His sketch told each student that at times they were sociable and at other times reserved. Ask yourself who that would not fit.
• Look up the sky. A free planetarium app shows which constellation the Sun is in on your birthday. Compare it with your sign.
• Use the four questions. For any claim: what was observed, what was inferred, does the evidence support it, and what is still open? Our guide to checking a viral science claim walks through it.

The verdict stays the same: tested fairly, astrology’s predictions do no better than chance. Enjoying it is a separate matter.