On 15 August 1977 the Big Ear radio telescope at Ohio State University was listening for signs of other intelligence. It could not steer; the Earth turned the sky through its beam, and any one point stayed in view for 72 seconds, no longer. In that window, close to the frequency that neutral hydrogen emits, it recorded a narrowband signal that rose, peaked and fell exactly as a steady distant source would. The chart below replays those 72 seconds in real time. When it is over, the signal is over. It was never heard again.
The chart sat in the stack for days. Jerry Ehman, a volunteer, spotted it; he circled the row and wrote “Wow!” in the margin. This copy has been waiting since. The window is the real one: 72 seconds, and the page will not hurry it.
The beam is crossing the source. The signal rises, peaks and falls as it passes, and the characters mark each ten-second sample of its strength.
The characters are not a message. Each is the signal’s strength in that ten-second sample, measured against the noise of the sky: 1 to 9 are the first nine multiples of the noise, and from 10 upward letters continue the count. U sits between 30 and 31. The signal peaked at about thirty times the background.
Ehman circled the reading and wrote “Wow!” beside it. He said later: “I mean, without thinking, I wrote ‘Wow!’. It was the most significant thing we had seen.”
The telescope was turned back on that part of the sky for the next month, and has returned to it since. The signal has not been recorded again.
The row is not a coded message. It is a measurement: the signal’s strength over six ten-second samples, in units of the background noise. 1 to 9 are the first nine multiples of the noise; from 10 upward, letters continue the count. U sits between 30 and 31, so the signal peaked at about thirty times the background, and it stayed in one column of fifty channels, which is why the row reads the way it does.
The signal came from the direction of Sagittarius. The telescope had two feed horns, and the recording cannot say which one heard the signal, so two positions are possible, about nineteen degrees south-east of the galactic plane; the declination is certain, −27°03′ ± 20′. No star at either position has been confirmed as the source.
None is confirmed. Interstellar scintillation, a weaker steady signal brightened briefly by the same twinkling that affects starlight, has been suggested; so has a source that sweeps in frequency like a lighthouse. A 2017 proposal blamed the hydrogen clouds of two comets, and astronomers dismissed it, because the comets were not in the beam at the time and do not emit strongly at that frequency. A 2024 preprint, from observations made at Arecibo in 2020, argues for a rare astrophysical event: a cold hydrogen cloud lit suddenly by a stellar flare. That is a candidate explanation, not a confirmed one.
The printout is preserved by the Ohio History Connection. The band around 1420 MHz is reserved for radio astronomy, and terrestrial transmissions in it are forbidden. The signal has been heard once, and once only. Ehman, who wrote that he resisted “drawing vast conclusions from half-vast data”, told an interviewer in 2019 that the signal certainly had the potential to be the first transmission from an extraterrestrial intelligence.
The recording was kept, so you may replay it. The event itself never repeated.
No transmission is involved. This page replays the recorded chart in your browser, and the window is real time: 72 seconds from first rise to last fall. The rise and fall are the beam crossing the source, and the page does not add anything to the record. The curve between the samples is this page’s own interpolation; the samples themselves are the published sequence 6EQUJ5.
Sources: Wikipedia, “Wow! signal” (fetched 4 August 2026); Big Ear Radio Observatory, “The Wow! Signal”. The page replays the record at its real length; it adds the interpolation, and says so. The intensity scale, the searches, the two positions and the preserved printout are all in the sources above.