The Wow! Signal: 72 Seconds From Sagittarius That Never Came Back

At 11:16 on a Monday night in August 1977, a radio telescope in a farm field in central Ohio heard something nobody has heard since. No one was there to notice. The telescope had no dish to steer and no operator on duty, and the computer in the small building beside it just did what it did every 12 seconds: it measured the static in 50 narrow radio channels and printed a column of digits and letters on a roll of fanfold paper. For about a minute that night, one of those columns climbed from ordinary noise to a value thirty times stronger and then sank back down. Days later a volunteer astronomer reading the printout at home circled the six characters in red ink and wrote one word in the margin: Wow!

That scribble became the most famous mark in the history of the search for extraterrestrial intelligence. The Wow! signal has been studied, re-measured and argued over for nearly half a century. It has been blamed on satellites, comets, military radios, a reflection off space junk and a natural flare in a cloud of hydrogen gas. It has been searched for, dozens of times, with some of the largest radio telescopes on Earth. It has never come back. And in the last two years a team working from photographs of the old printouts has found that the signal was stronger, a little higher in frequency and more precisely placed than anyone had realized.

This is the story of what Big Ear heard, why the six characters 6EQUJ5 still matter, and why the most careful explanation on the table today is stranger, in its own way, than little green men.

A frequency everyone would know

The Wow! signal makes sense only against an idea that was barely 18 years old when it arrived. In September 1959, two Cornell physicists, Giuseppe Cocconi and Philip Morrison, published a short paper in Nature with a modest title, “Searching for Interstellar Communications.” Their argument was simple. If another civilization wanted to be found by radio, it would have to pick a frequency that any curious species would already be listening to. The obvious choice was the natural radio note of hydrogen, the most common element in the universe.

That note comes from a tiny flip inside the hydrogen atom. The electron and the proton each spin, and when the electron’s spin flips from lined up with the proton’s to opposite it, the atom gives up a very small amount of energy as a radio wave at 1,420.405751768 megahertz, a wavelength of 21 centimeters. Any one atom does this only about once in ten million years, but space holds so much hydrogen that the galaxy hums with it. Harold Ewen and Edward Purcell at Harvard first detected that hum in 1951, and by the late 1950s radio astronomers everywhere were tuning to 1,420 MHz to map the spiral arms of the Milky Way.

Cocconi and Morrison reasoned that astronomers on another world would be doing the same thing, so a beacon parked near the hydrogen line would land in front of the very instruments most likely to catch it. The paper ends with a line that SETI researchers still quote: “The probability of success is difficult to estimate; but if we never search, the chance of success is zero.”

Someone took them up on it almost at once. In the spring of 1960, a young astronomer named Frank Drake pointed the new 85-foot Howard E. Tatel telescope at the National Radio Astronomy Observatory in Green Bank, West Virginia, toward two nearby Sun-like stars, Tau Ceti and Epsilon Eridani. He called it Project Ozma, after the princess of the land of Oz. Over about 150 hours he heard one false alarm, which turned out to come from an aircraft, and nothing else. The idea was born, though, and so was the habit of listening near 1,420 MHz.

Later researchers widened the target into a band they called the “water hole,” the quiet stretch of the radio spectrum between the hydrogen line and the lines of the hydroxyl molecule near 1,660 MHz. Hydrogen plus hydroxyl makes water, and the joke was that different species might gather there the way animals gather at a watering hole. The pun stuck because the physics behind it is real. Below about 1 gigahertz the galaxy’s own radio noise gets loud, and above about 10 gigahertz Earth’s atmosphere starts to get in the way, so the window around 1.4 to 1.7 gigahertz is about the quietest place to whisper across interstellar space.

A telescope you could walk on

The instrument that caught the Wow! signal did not look like anyone’s idea of a radio telescope. Big Ear, officially the Ohio State University Radio Observatory, was designed by John D. Kraus, an Ohio State electrical engineer best known for inventing the helical antenna. Instead of a dish, Kraus built what amounted to a giant funnel lying on the ground. At one end was a flat, tiltable reflector about 103 meters long and 33 meters tall. At the other end, across a ground plane of aluminum-covered concrete about the size of three football fields, was a fixed paraboloid that focused the incoming waves back toward a pair of horn antennas sitting between the two walls. Altogether it had the collecting power of a dish about 53 meters across.

Big Ear sat on Ohio Wesleyan University land beside the university’s Perkins Observatory, outside the town of Delaware, about 20 miles north of Columbus. Construction began in 1956, with much of the work done by Kraus and his students, and the telescope was switched on in 1963. The telescope could not turn. Its flat reflector could be tilted to pick a height above the horizon, and the Earth’s rotation then swept the sky past the beam like a slow searchlight in reverse.

From 1965 to 1971 Big Ear ran the Ohio Sky Survey, one of the great radio catalogs of the era, which listed roughly 20,000 radio sources, many of them faint galaxies and quasars seen for the first time. In 1972 the National Science Foundation cut its funding. Rather than let the telescope sit idle, Kraus and his colleague Robert Dixon turned it to the hunt for alien radio signals in 1973. With no money, the Ohio SETI program ran on volunteers, borrowed parts and stubbornness. It went on to become the longest-running SETI project in the world, a record noted by Guinness.

The machine that wrote it down

By 1977 the receiver was listening to 50 adjacent channels, each 10 kilohertz wide, a 500-kilohertz slice of spectrum centered near the hydrogen line. Every 12 seconds the system took a 10-second average of the power in each channel, spent 2 seconds crunching the numbers, and printed one line of results. The computer was an IBM 1130, a mid-1960s machine with a few tens of kilobytes of memory, connected to an IBM 1132 line printer that hammered out each line with an audible ka-chunk. The software that ran the search, a program called N50CH, had been written in their spare time by Dixon and a Big Ear volunteer named Jerry Ehman.

Space on the printout was scarce, so each channel’s reading was squeezed into one character. The scale measured how far the power stood above the average background, in units of the noise’s own typical swing, its standard deviation. A blank meant less than 1. The digits 1 through 9 meant just that. Anything 10 or higher got a letter: A for 10 to 11, B for 11 to 12, and on through the alphabet. Most of every page was blanks and 1s, with the occasional 2 or 3 as the noise jittered. A letter was rare. A letter late in the alphabet was something else.

The printouts piled up. Every few days a mechanical technician, Gene Mikesell, collected the paper and dropped it off for Ehman, who was helping the project as an unpaid volunteer. Ehman went through the pages at home, scanning for anything that stood out from the noise.

Six characters

The paper Ehman picked up around August 19 included the night of Monday, August 15. The line timestamped 22:16:01 Eastern Standard Time, which was 11:16 p.m. on the local daylight clock, showed nothing in most of its 50 columns. In channel 2, a stack of characters ran down the page: 6, E, Q, U, J, 5.

Decoded, that is a rise and fall. Six times background, then 14, then 26, then a peak between 30 and 31 that Ehman later put at about 30.5, then 19, then 5. It lasted six printed lines, 72 seconds. To a radio astronomer, the shape mattered as much as the strength. Because Big Ear was fixed, a point in the sky drifted through its beam in a predictable way, brightening as it neared the center and fading as it left, over about 72 seconds at that declination. A source that was really out among the stars, holding still while the Earth turned under it, should trace exactly that hill. 6EQUJ5 did. A nearby radio transmitter, a passing airplane or a glitch in the electronics would have had no reason to.

The signal also sat in a single 10-kilohertz channel, which meant it was narrowband. Natural radio emitters, from hot gas clouds to quasars, usually spread their energy across a wide range of frequencies, while transmitters built by engineers concentrate it. And it sat close to the hydrogen line, right where Cocconi and Morrison had said a beacon should be. Ehman circled the sequence with a red pen and wrote “Wow!” beside it. He later said it was the first thing that came to mind.

It is worth being clear about what the six characters are not. 6EQUJ5 is not a message, a code or a word. It is a record of loudness over time, rendered in a shorthand invented to fit on a narrow page. If something was encoded in the signal, Big Ear could not have seen it; any pattern faster than about 10 seconds or slower than 72 would have been averaged away.

A signal with two possible homes

Big Ear had one more quirk, and it has haunted the Wow! signal ever since. The receiver used two horn antennas side by side, aimed at slightly different points in the sky. A switch flipped between them dozens of times a second, and the system recorded the difference between the two. The trick canceled out much of the slow drift in the electronics, but it meant any real source in the sky should have passed through one horn’s beam and then, about three minutes later, through the other. The second horn would have recorded the same hill, upside down.

The second hill never came. The Wow! signal shows up in one beam only. Worse, the way the data were processed made it impossible to tell which horn had seen it. So the signal has two possible positions in the sky, separated by about two and a half minutes of right ascension, both at a declination of about 27 degrees south, in the constellation Sagittarius. One of them is often described as roughly 2.5 degrees south of the star group Chi Sagittarii. The nearest star easy to see by eye is Tau Sagittarii.

The missing second hill is the single strongest clue about what the Wow! signal was, and it points two ways at once. Either the source switched off, or faded, in the three minutes between beams, which would make it short-lived or intermittent. Or it was not a point in the sky at all and only seemed to be. Every hypothesis since 1977 has had to explain that silence.

There was a frequency puzzle too. When John Kraus summarized the detection for Carl Sagan in 1994, he gave a frequency of 1,420.3556 MHz. Ehman’s own detailed write-up, finished in 1998, put it at 1,420.4556 MHz, a tenth of a megahertz higher. The difference traced back to an ordering error years earlier: the university purchasing office had bought a local oscillator for the receiver at 1,450.5056 MHz instead of the 1,450.4056 MHz the engineers had specified. The software was written to compensate, and the two men had made the correction differently. Ehman’s figure, about 50 kilohertz above the hydrogen line, became the standard value for the next quarter century.

Looking again, and again

Ehman and his colleagues went back to the same patch of sky with Big Ear more than 50 times over the following months and years. They found nothing. “We should have seen it again when we looked for it 50 times,” he said in 1994. For a while he leaned toward the idea that a signal from the ground had bounced off a piece of orbiting debris and briefly mimicked a source in the stars, a worry that has only grown with the clutter described in our explainer on the Kessler syndrome. Later work showed how improbable that would be, since a reflector would have to be unusually large, shiny and well-placed, and stay in the beam just long enough to fake the hill. In a 1997 paper Ehman settled on a phrase that has become a kind of motto for the field: he refused to draw “vast conclusions from half-vast data.”

The most persistent searcher was not a professional astronomer at all. Robert H. Gray was a Chicago data analyst and SETI enthusiast who read about the Wow! signal and decided to keep looking for it. He started with a homemade receiver and a 12-foot dish in his backyard. In 1987 and 1989 he talked his way onto Harvard’s META system at Oak Ridge Observatory in Massachusetts and searched the Wow! coordinates, with no luck.

In 1995 and 1996, Gray and Kevin Marvel used the Very Large Array in New Mexico, a telescope far more sensitive than Big Ear, to stare at the Wow! region. They published their null result in the Astrophysical Journal in 2001. The VLA ruled out a steady source down to a level far below what Big Ear had seen. That mattered, because one popular explanation, interstellar scintillation, imagined a weak but steady signal that the turbulent gas between the stars had briefly magnified, the way air makes a star twinkle. The VLA should have caught such a source even in its dim phases, and it did not.

In 1995, H. Paul Shuch of the SETI League made drift scans of the area with a 12-meter dish at Green Bank and came up empty too. Gray went to Tasmania in 1999 and, with astronomer Simon Ellingsen, used the University of Tasmania’s 26-meter dish at Mount Pleasant to stare at the region for six 14-hour sessions, long enough to catch a beacon that switched on for a few minutes every few hours. Nothing. Gray told the whole story in a 2011 book, The Elusive Wow, and kept at it until near the end of his life. His last search, with the Allen Telescope Array in northern California, was published in 2020. He died in December 2021.

Before he died, Gray worked with the Columbia University astronomer David Kipping on a different kind of question. Suppose the Wow! signal came from a beacon that repeats, but irregularly, flashing on at random intervals the way a lighthouse might if its keeper were unpredictable. How often could such a beacon flash, and how long could each flash last, and still be consistent with one detection and decades of misses? Their paper, published in 2022 in the Monthly Notices of the Royal Astronomical Society, found that the idea cannot be ruled out, though the follow-ups make it uncomfortable. Before the extra observations, a repeater that fit the original detection had about a one-in-three chance of having been missed afterward. Including all the searches, that fell to under 2 percent. The model also boxed the length of each transmission between 72 seconds and about 77 minutes, and estimated that roughly two more months of continuous watching would be needed to push the case to a firm rejection.

In May 2022, Breakthrough Listen added another pair of null results. Its team used the Green Bank Telescope, the 100-meter dish that now dominates the valley where Project Ozma started, and the Allen Telescope Array to watch the region on the same night. The target was a Sun-like star that the Spanish amateur astronomer Alberto Caballero had flagged earlier that year: 2MASS 19281982-2640123, about 1,800 light-years away and the best-characterized Sun-like star inside one of the two Wow! beams. Neither telescope heard anything.

The explanations that didn’t hold

Over the decades the Wow! signal has drawn plenty of tidy solutions, and most have fallen apart under a closer look. The protected-band argument comes up first. Since 1959, international agreements have reserved the range around the hydrogen line for radio astronomy, so legal transmitters should not have been broadcasting there. That rules out intent, not accidents, and studies have documented illegal or leaking signals in the band. But a transmitter on the ground would have had to sit low on the southern horizon from Ohio, mimic a perfect beam pattern and show up in one horn but not the other.

In 2017 an astronomy professor at St. Petersburg College in Florida, Antonio Paris, proposed that two comets, 266P/Christensen and 335P/Gibbs, had been passing through the region in 1977 and that the hydrogen clouds around them could have produced the signal. The Big Ear team and other astronomers pushed back hard. The comets were not actually in either beam at the right moment, comets do not shine brightly at the hydrogen frequency, and there was no reason a comet would appear in one horn and not the other. The comet idea was dismissed.

Orbiting objects fare no better. A geostationary satellite sits too close to the celestial equator to reach a declination of 27 degrees south from Ohio, and a satellite in a high, looping orbit like the Soviet Molniya craft would have shot across Big Ear’s narrow beam in a fraction of a second instead of drifting through it for 72.

A second look at the paper

The most serious new work on the Wow! signal began not with a new telescope but with old paper. Between 2006 and 2017, a Big Ear volunteer named Marc Abel photographed more than 75,000 pages of Ohio SETI printouts, the whole run of N50CH output, before the paper could be lost. He shared copies with a handful of people, among them Russ Childers of the North American AstroPhysical Observatory, a group of former Big Ear volunteers who look after the telescope’s records.

Those photographs ended up with Abel Méndez, a planetary astrobiologist who runs the Planetary Habitability Laboratory at the University of Puerto Rico at Arecibo, and his project “Arecibo Wow!” Méndez’s team trained the open-source text recognition program Tesseract to read the 1970s printer’s characters, then checked its work with machine learning and human eyes. For the first time, the entire Ohio SETI record could be searched by computer instead of by one volunteer at a time, page by page.

The first paper from the project, posted in August 2024, began with observations Méndez had made in 2020 with the great Arecibo telescope, months before cable failures led to its collapse that December. In archived drift scans of the sky, the team found narrowband hydrogen-line signals coming from small, cold clouds of hydrogen gas. They looked like weak copies of the Wow! signal, about a hundred times fainter. That gave the group an idea.

Their hypothesis goes like this. Somewhere in that direction is a compact cloud of cold neutral hydrogen, colder than about 100 kelvins, because a warmer cloud’s atoms would move fast enough to smear the signal across more than 10 kilohertz. Normally such a cloud gives off a faint, steady hiss at 1,420 MHz. Then something hits it with a powerful jolt of radiation, perhaps a flare from a magnetar, a kind of neutron star with a magnetic field a thousand trillion times Earth’s. The jolt pushes the cloud’s atoms into a state in which they emit together rather than one at a time, amplifying the hydrogen line the way a laser amplifies light. Astronomers call the effect a maser, and the related quantum process, in which atoms release their energy in one coordinated flash, is called superradiance.

A flare like that could explain a lot at once. It would be narrowband and near the hydrogen line, because hydrogen is doing the emitting. It could be bright and brief, lasting a few minutes or less, which would let it show up in one horn and vanish before the other horn swung around. And it would almost never repeat from the same spot, because the trigger is rare and the cloud has to be lined up just so. Hydrogen masers built by engineers are the beating hearts of many of the world’s most precise clocks, the same family of timekeepers behind our look at how GPS finds you. But no one has ever caught a natural hydrogen-line maser in space, so the explanation stands or falls on finding another one.

The group’s closing note in that first paper had some fun with the alien alternative. If the Wow! signal really was a beacon, they observed, its makers might simply “deem it rude to direct a maser toward another entity.”

The second paper, posted on August 14, 2025, almost exactly 48 years after the detection, went back to the printouts with a finer comb. Its authors, including Kipping, worked out a mistake that had gone unnoticed for decades. The N50CH program printed the setting of the receiver’s second local oscillator, which shifted continuously to correct for the Earth’s motion, and the team found that the center value it was relative to had been 120.1 MHz, not 120.0, until December 13, 1977. Redoing the arithmetic moved the signal’s frequency up to 1,420.726 MHz, about 0.3 MHz above the hydrogen line rather than 0.05.

That small shift has a large meaning. If the source is hydrogen, the new frequency says it was moving toward the Sun at about 84 kilometers per second, a speed that does not fit the ordinary rotation of the galaxy in that direction but is not impossible for a cloud or a star system with its own motion. The team found at least three stars in the European Space Agency’s Gaia catalog in the right region with similar velocities.

The rest of the revisions sharpened the picture. A careful fit of the printed values gave a peak of 30.1 standard deviations and a crossing time of 73.4 seconds. Recalibrating against known radio sources recorded on the same printouts put the flux density at about 256 janskys, plus or minus 63, compared with older estimates of 54 or 212. That is far brighter than almost anything in the radio sky at that frequency. The source had to be smaller than about 2 arcminutes across. And the two possible positions moved by about 7 arcminutes and shrank to patches more than three times narrower than before, though the declination still carries a 20-arcminute uncertainty from Big Ear’s tall beam.

The team also hunted for anything mundane. They searched newspapers and records from the months around the detection for local transmitters and found none that fit. They checked whether the signal could be the second harmonic, the doubled frequency, of a strong broadcast near 710 MHz, the range of UHF television channels 53 and 54, and it did not fit. They looked at the Sun, which on August 15, 1977 was near the quiet bottom of its cycle with no major flares. Their conclusion was that nothing in the data pointed to interference and that the evidence lined up with an astrophysical source.

Then there were the other circles. Scanning the transcribed record, Méndez’s group found narrowband signals from January 1978 that the Ohio SETI team had also circled by hand. Nicknamed Wow2 and Wow3, they had the right beam shape but were much weaker and came from different parts of the sky. If the hydrogen-flare idea is correct, the Wow! signal may simply have been the brightest of a family of events Big Ear was catching all along, unnoticed, because nobody had a computer to read the pages.

What Dixon predicted

It is tempting to see the hydrogen-cloud idea as a letdown. It isn’t, and the people who ran Big Ear said so long ago. Robert Dixon, the program’s director, put it plainly in 1985: “We’ve always felt we may not discover intelligent life, but we will discover a new natural phenomenon.” Carl Sagan made a similar point in Pale Blue Dot in 1994, noting that unusual gas clouds might produce narrow signals that would fool searchers.

Astronomy has been here before. In 1967 a graduate student at Cambridge, Jocelyn Bell, noticed a strange, regular pulse in her chart recordings. Her group half-jokingly labeled it LGM-1, for “little green men.” It turned out to be the first pulsar, a spinning neutron star, and a Nobel Prize followed. Radio astronomers learned to check for natural explanations first, and the Wow! signal has always been treated the same way. The point is less to find aliens than to find out what made those six characters.

The search for meaning in strange signals from the sky has a long and sometimes silly history. In 1835 a New York newspaper persuaded a large share of its readers that bat-winged people lived on the Moon, as told in our story of the Great Moon Hoax. The Wow! signal belongs to the opposite tradition: a real measurement, cautiously reported, argued over in journals and never oversold by the man who found it. In a 2019 interview, Ehman said only that “the Wow! signal certainly has the potential of being the first signal from extraterrestrial intelligence.” He has rarely gone further than that.

The telescope is gone; the question isn’t

Big Ear itself did not survive the argument. The university sold the land to developers, and despite protests from astronomers and a campaign by volunteers, the telescope was dismantled in 1998 to make room for an expanded golf course and new homes. The full Ohio SETI search had run from 1973 to 1995, and a last low-budget survey, recorded digitally, carried on for a few more years. The Perkins Observatory still stands nearby.

The paper survived, though, and so did the people who looked after it. Big Ear veterans helped the Arecibo Wow! team untangle how the old receiver and software worked. The Planetary Habitability Laboratory’s project page, updated in September 2026, now gives an even higher flux estimate, more than 300 janskys, and slightly refined positions near 19h25m26s and 19h28m18s of right ascension. A third paper, Arecibo Wow! III, is planned for August 2027, the signal’s 50th anniversary.

There is a practical program behind it. If the source was a flaring hydrogen cloud, there should be more such clouds and more flares, and modern telescopes with wide fields and fine frequency resolution should eventually catch one in the act. If a flare is seen and it looks like 6EQUJ5, the Wow! signal will most likely have an answer, and a natural one. If decades pass and nothing like it shows up, the field will be left where Ehman left it in 1998, with one strong, clean, unrepeated measurement and no good way to judge it.

Meanwhile people have tried answering it anyway. On the signal’s 35th anniversary in 2012, the Arecibo Observatory beamed a transmission of about 10,000 messages collected on Twitter, as part of a promotion for a television series, as a symbolic reply. The former observatory director who oversaw it later said the transmitter overheated shortly after the first transmission began. None of the stars targeted was actually inside the area where the signal came from.

That may be the best summary of the Wow! signal: half a century of careful, passionate effort aimed at a point in the sky that was never quite where anyone thought. The new frequency, the narrowed positions and the 75,000 photographed pages have given the hunt its best map yet. Whatever lit up channel 2 for 72 seconds in 1977, it is still out there, or it was a flash that is gone forever. Either way, someone heard it, and someone wrote it down.

Further reading