The Drake equation, term by term: what we've measured and what's still a guess

In November 1961 a young astronomer, Frank Drake, hosted the first scientific meeting on the search for extraterrestrial intelligence (SETI) at the radio observatory in Green Bank, West Virginia. A year earlier he had pointed a radio dish there at two nearby Sun-like stars in an experiment called Project Ozma, and heard nothing. To give the meeting an agenda, he wrote down an equation. The attendees included a young Carl Sagan and physicist Philip Morrison, and they called themselves "The Order of the Dolphin," after dolphin researcher John Lilly, who was among them.

Drake wasn't trying to produce a precise number. The equation organized the discussion around what we would need to know. The group's guesses were optimistic: one new star a year, a fifth to a half of stars with planets, one to five suitable worlds per system, life and intelligence wherever possible, and 10% to 20% of civilizations communicating. Their conclusion was that the number of civilizations roughly equals how many years each one survives: somewhere between 1,000 and 100 million.

N = R* × fp × ne × fl × fi × fc × L. Each symbol is one question:

**Star formation rate (R*).** A 2015 meta-analysis found the Milky Way turns about 1.65 solar masses of gas into stars each year. Most new stars are smaller than the Sun, so that means a few stars a year, not far from the 1961 guess.

Stars with planets (fp). A guess in 1961, close to a certainty now. A gravitational microlensing study published in Nature in 2012 found that Milky Way stars host, on average, at least one planet each. As of mid-October 2026, NASA's Exoplanet Archive lists 6,451 confirmed planets around 4,847 stars. The Kepler space telescope (2009 to 2018) found 2,787 of them, plus 550 in its follow-on K2 mission. TESS, launched in 2018 to survey nearly the whole sky, has added 1,012 so far, with thousands of its candidates still awaiting confirmation.

Worlds that could support life (ne). Here the data thins out. A 2021 analysis of Kepler data led by Steve Bryson of NASA Ames estimated that a Sun-like star has, on average, 0.37 to 0.60 rocky planets in its habitable zone, where liquid water is possible, and up to 0.88 under a looser definition. The error bars are wide because so few such planets have been found. Being in the habitable zone doesn't guarantee life either: Venus and Mars sit near the edges of ours. The nearest known example is Proxima Centauri b, about 4.2 light-years away, orbiting a red dwarf prone to violent flares.

Life (fl). We have exactly one example. Life appeared on Earth early, more than 3.5 billion years ago, and a 2020 analysis by Columbia University astronomer David Kipping found that this favors, by odds of at least 2.8 to 1, life arising quickly where conditions allow. But no life has been found beyond Earth: NASA's September 2025 Mars rock result is a "potential biosignature," and the dimethyl sulfide signal on K2-18b is disputed.

Intelligence (fi) and technology (fc). Billions of species have lived on Earth and only one built a radio, after about four billion years. The same Kipping analysis leans slightly, at 3 to 2, toward intelligence being rare. And we have no idea whether other civilizations would transmit at all, prefer lasers, or choose silence.

Lifetime (L). We have been broadcasting for about a century, and that is our only data point. Carl Sagan thought this was the decisive factor: if civilizations tend to destroy themselves quickly, the galaxy is quiet even if life is common. That is why the 1961 meeting concluded that N roughly equals L. Once the other factors roughly cancel out, the question becomes how long civilizations last.

To see how open the answer is, here are two calculations of our own, not any scientist's estimate. Both use the same measured first three terms: two new stars a year, planets around all of them, and half a suitable world per system.

Serious studies span a similar range. Oxford's Anders Sandberg, Eric Drexler and Toby Ord showed in 2018 that once realistic uncertainties are included, there is a high chance we are alone in the galaxy. Tom Westby and Christopher Conselice of the University of Nottingham found in 2020 at least 36 civilizations in the galaxy, assuming each communicates for 100 years, with the nearest probably up to about 17,000 light-years away.

Adam Frank and Woodruff Sullivan calculated in 2016 that unless the odds of a habitable-zone planet producing a technological species are below about one in 10 to the 24th power, we are not the first technological civilization in cosmic history. Note that this answers a different question: not how many civilizations exist in the galaxy now, but whether there has ever been another one, anywhere.

The Drake equation is not a measurement and not a proof. It counts only civilizations in our galaxy, only those giving off signs now, and only those we could detect. It ignores civilizations that spread from star to star, which is exactly where the Fermi paradox comes in: if even one did, it should be everywhere by now. And it says nothing about UFOs, because it doesn't deal with visits to Earth.

Over the years it has spawned variants. In 2013 MIT astronomer Sara Seager proposed an equation that counts planets where we could detect gases given off by living things, a more realistic target for the telescopes of the coming decades. The Pentagon took an interest too: "An Introduction to the Statistical Drake Equation," written in 2010 under the Defense Intelligence Agency's AAWSAP program, was published in the sixth batch of UFO files in September 2026. Read the document here, and more on the program itself.

Three kinds of discovery could turn the equation from a guess into science. A microbe on Mars, or on one of the icy moons of Jupiter or Saturn, that arose independently of life here would show that fl is high. A clear biological gas in a distant planet's atmosphere, a job for the James Webb Space Telescope and its successors, would show the same on a galactic scale. And a single confirmed technosignature, unlike every candidate so far, would prove N is greater than one. How that would be verified and announced is in our first contact protocol explainer.