How Do We Search for Extraterrestrial Intelligence? SETI Explained

On August 15, 1977, the Big Ear radio telescope at Ohio State University recorded a narrowband burst of radio energy that lasted 72 seconds and peaked at about 30 times the background noise. Days later, astronomer Jerry Ehman spotted it on a printout and wrote "Wow!" in the margin. It has never been heard again. The search for extraterrestrial intelligence (SETI) is the scientific effort to find signals like it, and this article explains how that search works, what has been checked, and why more than 65 years of listening has produced silence.

What SETI looks for, and what it does not

SETI does not search for microbes or fossils. That work belongs to astrobiology, which looks for biosignatures, chemical traces of life such as gases in a planet's atmosphere (see How Do We Detect Exoplanet Atmospheres, and What Are Biosignatures?). SETI looks for technosignatures, which the field defines as any measurable property or effect that provides scientific evidence of past or present technology. A radio beam, a laser flash, an industrial pollutant, and a structure built around a star all count.

The logic rests on a modest assumption: if another technological civilization exists and uses physics we also understand, some of its activity might be detectable across interstellar distances. Technology tends to be unnatural in recognizable ways. It is narrower, brighter, more regular, or more concentrated than anything a star, planet, or gas cloud produces by itself.

That makes SETI a search for contrast. Nature produces broad, smooth, noisy radio emission. A transmitter produces a spike at one frequency. Nature does not usually flash on a nanosecond timescale from a single point. Engineering does. The entire method comes down to building instruments sensitive enough to notice such contrasts, then rigorously ruling out the human-made and natural impostors.

Why the question is scientific, not speculative

The search is testable even though the answer is unknown. Each survey specifies a set of stars, a frequency range, a sensitivity, and a duration, and it reports what it excluded. A null result is real information: it says that no transmitters of a given power existed in that region of parameter space during that time. This is why SETI results are often published as limits, for example "no transmitters stronger than X from these 252 stars," rather than as detections.

A short history, from a 1959 paper to Breakthrough Listen

The modern search starts with a short published argument. In 1959, Giuseppe Cocconi and Philip Morrison proposed that the microwave part of the spectrum was a sensible place to look for signals from other stars. The next year, a young astronomer acted on that idea.

Year	Milestone	What happened
1959	Cocconi and Morrison paper	First proposal to search the microwave spectrum for interstellar signals
1960	Project Ozma	Frank Drake used the 26-meter (85-foot) telescope at Green Bank, West Virginia, to observe Tau Ceti and Epsilon Eridani near 1,420 MHz
1961	Green Bank meeting	Drake presented his equation to a small group of researchers, including Carl Sagan and Philip Morrison
1971	Project Cyclops	A NASA-funded study sketched a 1,500-dish array costing about US$10 billion; it was never built
1974	Arecibo message	A largely symbolic message was transmitted toward the globular cluster M13, about 25,000 light-years away
1977	Wow! signal	Big Ear recorded the strongest candidate signal in the field's history
1980s-1999	Sentinel, META, BETA	Harvard-led surveys used digital spectrum analyzers with rising channel counts; BETA ran from 1995 until wind damage ended it in 1999
1992-1993	NASA program	A federally funded SETI program began, and Congress canceled it about a year later
1995	Project Phoenix	The SETI Institute resumed targeted searches with private funds, running until about 2015
1999	SETI@home	Volunteer computing project released to the public on May 17, 1999
2007	Allen Telescope Array	First 42 dishes became operational
2015	Breakthrough Listen	Announced July 20, 2015, with $100 million over 10 years

Project Ozma deserves a closer look because it sets the template. Drake observed for roughly 150 hours, intermittently, over about four months, scanning a 400-kilohertz band with a receiver bandwidth of 100 hertz. Early on, a signal appeared on April 8, 1960. It turned out to come from a high-flying aircraft. That false alarm, in the very first search, previewed the central difficulty of the field: human technology is everywhere, and it looks a lot like what you are trying to find.

The funding story is also part of the history. Public money for SETI has been unstable. Funding was briefly cut in 1981 and restored in 1982, and the NASA program of the early 1990s was canceled after about a year. Much of the work since has been privately funded, from the SETI Institute's Project Phoenix to the $100 million Breakthrough Listen initiative led by Yuri Milner's Breakthrough Initiatives.

Why radio, and why near 1,420 MHz

Radio waves have three practical advantages for interstellar signaling. They travel through the gas and dust of the galaxy with little absorption. They are cheap to generate and detect compared with other wavelengths. And, as Cocconi and Morrison argued, there are natural focal points in the spectrum that two civilizations with no prior contact might both choose.

The most famous is the hydrogen line, near 1,420 MHz (21 centimeters). Neutral hydrogen is the most abundant atom in the universe, and its emission at this frequency is one of the first things any radio astronomer learns to detect. A civilization that studies the sky in radio would know it. The idea is like two strangers who have never spoken agreeing to meet at the central clock in a train station: the point is not that the clock is special, but that both parties can reason their way to it.

The water hole

A second suggestion is the water hole, the band between 1,420 and 1,662 MHz. It lies between the hydrogen line and the strongest line of the hydroxyl radical (OH). Hydrogen and hydroxyl are the two pieces of a water molecule, and the interstellar gas that contains them absorbs radio noise at those frequencies, leaving the gap between them comparatively quiet. Bernard Oliver coined the name in 1971, as a pun on a watering hole where species gather. If the idea is right, the water hole is a natural meeting place. If it is wrong, it is merely a reasonable place to start.

Narrowband: the signature of a transmitter

Almost everything a radio telescope detects from the cosmos is broadband. Stars, gas clouds, pulsars, and galaxies emit energy smeared across many frequencies. A transmitter, by contrast, concentrates its power into a very narrow slice of spectrum, because that is the most efficient way to make a signal stand out. The Wow! signal, for example, had a bandwidth under 10 kilohertz, confined to a single receiver channel.

Real SETI searches therefore scan for narrow spikes, often with channels only a few hertz wide. A spike is not enough on its own. The signal must also be at a plausible frequency, appear when the telescope points at the target, vanish when it points away, and show a drift in frequency consistent with the motion of a distant planet. A transmitter on a rotating, orbiting planet produces a changing Doppler shift, the same effect that changes the pitch of a passing siren. A signal that does not drift in the expected way, or that appears regardless of where the telescope points, is probably local.

The interference problem: why almost every candidate fails

Most of the effort in SETI goes into rejecting false positives. The Earth is saturated with radio transmitters, including phones, satellites, radar, aircraft, power equipment, and the telescope's own electronics. This radio frequency interference (RFI) can mimic the narrowband spikes the search is designed to find.

A few well-known candidates illustrate how the process works.

Candidate	When	What was seen	Outcome
Aircraft signal in Project Ozma	April 8, 1960	A signal during the first-ever search	Traced to a high-flying aircraft
Wow! signal	August 15, 1977	72-second narrowband burst near 1,420 MHz	Never recurred; unexplained
SHGb02+14a	Announced 2004	Spikes in SETI@home data	Most significant SETI@home candidate; never confirmed as extraterrestrial
Breakthrough Listen first results	April 2017	Eleven significant events among 692 nearby stars at 1.1-1.9 GHz	All judged to be RFI
BLC1	Observed April-May 2019; reported December 2020	Narrowband signal at 982.002 MHz toward Proxima Centauri	Studies in October 2021 concluded it resembled terrestrial interference

The BLC1 case in detail

BLC1 ("Breakthrough Listen Candidate 1") is the best modern teaching example. The signal was found in about 30 hours of Parkes Observatory data in Australia, collected in April and May 2019. It sat at 982.002 MHz, and the telescope's beam, about 16 arcminutes wide, contained Proxima Centauri, the nearest star to the Sun, which hosts a known planet. The frequency changed over time in a way that looked at first like a Doppler drift.

Those details made it look serious. But the drift increased where Earth's rotation alone would have made it decrease, and it did not match the motion of Proxima b either. Repeat observations did not find the signal again. On October 25, 2021, two studies concluded it was unlikely to be a technosignature because it resembled previously detected terrestrial interference. A good candidate had been run through a good process and rejected. That is the system working, not failing.

How to tell the sky from the ground

The standard trick is to nod the telescope. Observers point at the target, then at a nearby patch of sky, then back, repeating the cycle. A real source in the target's direction appears only when the telescope looks at it. Interference that leaks in from the side or from the ground tends to appear in every pointing, or in an inconsistent pattern. Combine that with frequency-drift analysis, polarization checks, and independent follow-up at a second telescope, and most false alarms fall away quickly. The ones that do not are the rare, interesting ones.

The Wow! signal: what we know

The Wow! signal remains the field's most discussed anomaly, so it is worth separating the facts from the folklore. The details below come from the published record as summarized in the sources at the end of this article. A dedicated account is in The Wow Signal 1977.

• What was recorded. A narrowband signal at about 1,420.46 MHz, roughly 50 kilohertz above the nominal hydrogen line, with a bandwidth under 10 kilohertz.
• What "6EQUJ5" means. The string Ehman circled is not a message. It is a record of signal intensity over time, in a code where each character stands for a range of signal-to-noise. The highest value, "U," corresponds to about 30 standard deviations above background.
• Why 72 seconds. That is how long a fixed point on the sky stays in Big Ear's beam as Earth rotates. The rise and fall in intensity matches what a distant, fixed source would produce.
• Where it came from. The direction lies in the constellation Sagittarius, with some ambiguity because the telescope had two feed horns. A 2022 paper identified three candidate Sun-like stars within the possible coordinates.
• What has not happened. The signal has not recurred. Ehman's own follow-up searches failed, as did later efforts with the META project, the Very Large Array, and a targeted Breakthrough Listen search on May 21, 2022, which found no technosignature candidates.

Proposed explanations

Several ideas have been floated, and none has won consensus. Ehman once suggested a reflection from space debris in Earth orbit, then softened that view after finding the required conditions unrealistic. In 2017, Antonio Paris proposed that hydrogen clouds around two comets could produce the signal, but astronomers dismissed the idea because the comets were not in the beam at the right time. A 2024 preprint proposed that stellar emission energized a cold hydrogen cloud and caused a sudden brightening.

The signal's frustrating feature is that it was seen once, by one telescope. Science treats single, unrepeatable events cautiously, because there is no way to test them. That is why the Wow! signal is a famous puzzle rather than a discovery.

The cosmic haystack: why silence is weak evidence

Not finding a signal in more than 65 years sounds damning. It is less so than it appears, because the space of possible signals is enormous. Researchers describe it as a multi-dimensional haystack. To confirm a detection, a survey has to match the signal on at least six axes at once:

1. Direction: which of billions of stars, or which patch of sky.
2. Frequency: which part of a radio spectrum thousands of times wider than any one channel.
3. Time: the signal might be continuous, periodic, or a rare burst.
4. Sensitivity: whether the telescope is powerful enough to hear it at that distance.
5. Polarization and modulation: whether the receiver is set up to decode its structure.
6. Technology: whether the civilization uses radio at all.

A survey that covers a tiny fraction of each axis can truthfully claim to have searched, while covering a vanishing fraction of the whole space. Imagine searching an ocean for fish by dipping a bucket: finding none in one scoop says almost nothing about the ocean.

Breakthrough Listen's design shows how the field tries to cover more of the space. Operating from January 2016, it uses the Green Bank Telescope in the north and the Parkes Telescope in the south for radio, and the 2.4-meter Automated Planet Finder at Lick Observatory for optical laser searches. Its stated targets are one million nearby stars, the centers of at least 100 nearby galaxies, and the plane and center of the Milky Way, plus exotic objects such as white dwarfs, neutron stars, and black holes. A 2020 analysis of 252 nearby stars at 3.95-8.00 GHz found no transmitters comparable to those at Arecibo, a limit that says something concrete about how common such transmitters are. For the other side of the sky, see the ranking of the largest radio telescope arrays, and for how the dishes work, see How Do Radio Telescopes Work?.

The Drake equation: a framework for ignorance

In 1961, Frank Drake wrote down the equation that organizes the question. It estimates N, the number of communicative civilizations in the Milky Way, as the product of seven terms.

Term	Meaning	How well known?
R*	Rate of formation of suitable stars	Reasonably well constrained by astronomy
fp	Fraction of stars with planets	A 2012 microlensing analysis suggested it may approach 1
ne	Number of life-supporting planets per system	Kepler-based estimates in 2013 suggested billions of Earth-sized planets in habitable zones
fl	Fraction of those planets where life appears	Unknown: one example
fi	Fraction where intelligence emerges	Unknown
fc	Fraction that develop detectable technology	Unknown
L	How long such civilizations emit detectable signals	The hardest term; little basis for a value

At the 1961 meeting, the attendees' guesses produced values of N from about 20 to 50 million. Later estimates in the literature span many orders of magnitude, from values far below one to millions, which shows how much the answer depends on terms nobody can measure.

Critics point out that the equation is not a prediction. Its first three terms now rest on astronomical data, but the biological and social terms are conjecture, so the output can land far below one or far above it. Defenders reply that Drake designed it as an agenda for discussion, not a calculator, and it has served that purpose for six decades. The full treatment is in The Drake Equation: How Many Aliens Exist?, and the puzzle of the silence itself is the subject of The Fermi Paradox and What Is the Great Filter?. Which stars are the best targets depends partly on the habitable zone and on the exoplanets we have found.

Optical SETI: looking for flashes

Radio is not the only option. In 1961, R. N. Schwartz and Charles Townes, who helped invent the laser, published a paper in Nature proposing that interstellar communication could use optical masers. NASA's 1971 Cyclops study discounted the idea, reasoning that a laser able to outshine its host star would be too hard to build. Later paper studies argued otherwise: a high-energy infrared laser focused by a 10-meter mirror could appear thousands of times brighter than the Sun to an observer in the beam's path.

The case for lasers has trade-offs. A laser carries a lot of information and, because the beam is narrow, cannot be intercepted by accident. The same narrowness means it must be aimed at Earth to be seen, and a pulse might arrive for a fraction of a second. Searches therefore watch many stars for short flashes.

Several programs do this today:

• A Harvard-Smithsonian laser detector on a 155-centimeter telescope surveyed about 2,500 stars from October 1998 to November 1999 and found no sign of an intentional laser signal.
• The Breakthrough Listen optical program uses the 2.4-meter Automated Planet Finder at Lick Observatory to look for continuous laser lines in the spectra collected during planet hunts.
• The SETI Institute's Laser SETI uses cameras that survey the night sky for millisecond laser pulses.
• PANOSETI installed two telescopes at Lick Observatory in January 2020 for wide-field searches for pulsed near-infrared signals.

No intentional optical signal has been confirmed. A 2017 study of Tabby's Star that looked for possible laser emission also found no technology-related signals.

Technosignatures beyond signals

Modern SETI has widened its target list. Instead of waiting for a message, researchers ask what a technological civilization would do to its planet or star that we could see.

Heat, pollution, lights, and megastructures

• Infrared excess. Freeman Dyson proposed in 1960 that a civilization could capture its star's energy with a giant shell or swarm of collectors. Such a structure would absorb starlight and re-emit it as waste heat, making the star look unusually bright in the infrared. Fermilab has run a survey of infrared satellite data since 2005, finding 17 "ambiguous" candidates, none confirmed. The concept is explained in What Is a Dyson Sphere?, and the scale of energy involved ties into the Kardashev scale.
• Industrial gases. Humans emit nitrogen dioxide and chlorofluorocarbons, which can be detected from space. Similar gases in an exoplanet's atmosphere might signal industry. The complication is that natural chemistry can mimic pollution, as Titan's haze shows.
• City lights. Astronomers have proposed detecting artificial lighting on exoplanets. One analysis found that sodium street lights could be told apart from natural sources by next-generation telescopes.
• Megastructure transits. A 2005 paper proposed that large artificial objects would leave distinctive shapes in a star's light curve as they pass in front of it.
• Artifacts. Some researchers argue that probes or other objects within the solar system are a well-justified target, according to a 2021 review of mission concepts sponsored by NASA.

NASA has helped formalize the field. A 2018 NASA Technosignatures Workshop produced a report published in January 2019, and in June 2020 the agency awarded its first SETI-specific grant in about three decades, funding a non-radio technosignature search and an online library of technosignatures.

Tabby's Star: a lesson in restraint

The best-known "megastructure" scare was KIC 8462852, also called Tabby's Star or Boyajian's Star, a star about 1,470 light-years away in Cygnus. Citizen scientists in the Planet Hunters project noticed irregular dips in its brightness in data from NASA's Kepler telescope. The largest dips reached about 15 percent in 2011 and about 22 percent in 2013, far deeper and less regular than a planet would cause.

An alien structure was one proposal. It was not the favored one. A 2018 study found that the material blocking the light filters different wavelengths by different amounts, which points to dust rather than an opaque object. Other ideas include comet swarms, intrinsic changes in the star, disrupted exomoons, and a planet consumed by its star. No hypothesis fully explains the light curve, and an artificial structure is disfavored but considered testable. The episode shows how the field is supposed to behave: take an anomaly seriously enough to test it, and be willing to let the more mundane explanation win.

Machine learning joins the hunt

Searches now generate more data than people can review by hand, so researchers use algorithms to flag unusual signals. A deep-learning study of 820 nearby stars flagged eight possible technosignatures for follow-up. As of the sources consulted here, none has been confirmed, and flagged signals have a long history of being traced to interference. Breakthrough Listen has also analyzed data on the interstellar object 3I/ATLAS in 2025. According to a summary of the results, about two million detections were filtered down to 211 signals of interest, none indicating a non-natural origin.

SETI@home: volunteer computers listen

Analyzing radio data takes computing power, and in 1999 the SETI@home project found a creative answer: ask the public to donate it. Volunteers installed a screensaver that processed small chunks of data from the Arecibo radio telescope, and later from the Green Bank Telescope, when their computers were idle. The data were collected "piggyback," while the telescopes did other science, and because Arecibo lacked a broadband connection at the time, data were physically mailed on hard drives to Berkeley.

The project released its software to the public on May 17, 1999 and drew millions of participants worldwide, among the largest volunteer computing efforts. On March 31, 2020, UC Berkeley stopped sending new data for processing, citing diminishing returns and a need to analyze the results already collected. That December, the Arecibo Observatory that had supplied much of the data collapsed. SETI@home never confirmed an extraterrestrial signal. It did identify several candidate sky positions for further study, and it proved that volunteer computing is a workable approach to large-scale signal analysis.

What detection would mean, and why it would not be a conversation

If a promising signal appeared, the response would be procedural before it was philosophical. The International Academy of Astronautics established a SETI Permanent Study Group and a post-detection taskgroup in 2005. A scale for rating the significance of a claimed detection, the Rio scale, was proposed in 2000 and revised in 2018. The central principle is verification: independent teams on independent instruments must confirm the signal before any announcement, because a premature claim would likely collapse.

The physics of distance shapes everything afterward. If a signal arrived from a source 1,000 light-years away, a reply would take 1,000 years to get there, and any answer another 1,000 years to return. A single round trip would span the entire history of written civilization several times over. A confirmed SETI detection would be a message in a bottle, not the start of a dialogue. For perspective on the distances, see How Long Would It Take to Travel to the Nearest Star?.

There is also a separate, contested question: whether humanity should transmit rather than listen. That field is sometimes called METI (messaging to extraterrestrial intelligence), and the 1974 Arecibo message to M13 is its best-known example. SETI proper is passive, a receiving effort, and it does not broadcast.

Common confusions

Is SETI looking for aliens or for signals?

Signals. SETI cannot detect a living organism across interstellar distances. It looks for evidence that technology exists, because technology can be bright, loud, or unusual enough to cross the gap. Searches for the chemistry of life on planets are a related but different field, covered in Astrobiology: The Science of Life Beyond Earth and The Search for Extraterrestrial Life: Where Do We Stand?.

Did the Wow! signal prove aliens exist?

No. It is an unexplained, unrepeated event, and science needs repetition to draw conclusions. Its intensity, its bandwidth, and its apparent origin on the sky make it intriguing, but it has no confirmed explanation either way.

Does silence prove we are alone?

No. Silence so far is consistent with several scenarios, including that technological life is rare, that it is short-lived, that it does not use the channels we search, or that our searches have covered only a tiny part of the haystack. The Fermi paradox articles linked above explore these possibilities.

Are mysterious radio bursts from space evidence of aliens?

Not by default. Natural sources such as pulsars and fast radio bursts produce intense, short radio signals, and natural explanations remain the default for such events. Scientists consider artificial origin only after natural and terrestrial explanations are exhausted. Between stars lies the interstellar medium, which can distort signals in ways that complicate this analysis, and the chemistry of astrochemistry explains many spectral features that once looked puzzling.

Frequently asked questions

Q: What does SETI stand for, and what does it do? A: SETI stands for the Search for Extraterrestrial Intelligence. It uses radio telescopes, optical telescopes, and data analysis to look for technosignatures, evidence of technology from other star systems. It does not search for simple life; that is astrobiology. No confirmed signal has been found.

Q: Why do SETI searches focus on 1,420 MHz? A: That is the frequency of the hydrogen line, from the most abundant atom in the universe, so any radio-capable civilization would know it. The surrounding "water hole," between 1,420 and 1,662 MHz, is also comparatively quiet because interstellar hydrogen and hydroxyl absorb radio noise there. It is a reasoned guess, not a proven channel.

Q: What is the Wow! signal? A: It is a 72-second narrowband radio signal recorded by Ohio State University's Big Ear telescope on August 15, 1977, at about 1,420 MHz and roughly 30 standard deviations above background. It has never recurred, and no explanation has gained consensus. Follow-up searches, including one in 2022, found nothing.

Q: Has SETI ever found a real alien signal? A: No. Candidates such as the Wow! signal, SHGb02+14a, and BLC1 have been examined and either remain unexplained, were never confirmed, or were attributed to terrestrial interference. A confirmed detection would require independent verification by other teams and instruments.

Q: How much money goes into SETI? A: Public funding has been unstable. NASA's 1990s program was canceled by Congress about a year after it began. Breakthrough Listen, announced in 2015, committed $100 million over 10 years from private funding. The SETI Institute's earlier Project Phoenix was also privately funded.

Q: Can I help search for aliens? A: SETI@home let volunteers process radio data from 1999 until it stopped sending new work on March 31, 2020. Breakthrough Listen publishes data through an open archive. Citizen-science projects such as Planet Hunters have also contributed, as in the discovery of Tabby's Star.

The takeaway

SETI is best understood as a long exercise in careful negatives. Every survey narrows the space of possibilities a little: no strong narrowband transmitters from these stars, no pulsed lasers from those, no excess infrared from that sample. The famous anomalies, the Wow! signal and BLC1, are valuable mostly for what they teach about method. One was seen once and cannot be tested; the other was tested and found to be ours. Because the search has covered so little of the haystack, its silence says far less than people assume, and because the method is sound, a real detection would be recognizable. Whether that day comes in a decade or never, the instruments built for the search, from survey arrays to laser detectors, are also quietly mapping a universe whose natural behavior is stranger than we thought.

Sources

• Wikipedia: Wow! signal
• Wikipedia: Project Ozma
• Wikipedia: Breakthrough Listen
• Wikipedia: BLC1
• Wikipedia: SETI@home
• Wikipedia: Search for extraterrestrial intelligence
• Wikipedia: Optical SETI
• Wikipedia: Technosignature
• Wikipedia: Water hole (radio)
• Wikipedia: Drake equation
• Wikipedia: KIC 8462852
• SETI Institute: The Drake equation