The number has a birthday: October 2016

“Two trillion galaxies” entered popular science on a specific date. In October 2016, a team led by Christopher Conselice at the University of Nottingham published The Evolution of Galaxy Number Density in The Astrophysical Journal. They did not photograph two trillion galaxies. They took the galaxies we can see, modelled how the population thinned out at fainter and earlier epochs, and integrated the result across cosmic time.

Two findings drove the headline. First, the model implied that roughly 90% of galaxies are too faint for current telescopes to detect at all. Second — and this is the part the headline always drops — the early universe held far more small galaxies that have since merged into bigger ones. So part of the two trillion is a tally of galaxies that no longer exist as separate objects. It is less a census of the universe today than a sum over its whole history.

How you actually count galaxies

You cannot photograph the whole sky deeply enough to count galaxies one by one — it would take lifetimes. So astronomers take a “core sample.” Point the telescope at a tiny, deliberately empty-looking patch, expose for days, count everything in frame, then multiply by how many such patches it would take to tile the entire sky. That is exactly what the widget above does.

The landmark samples: the original Hubble Deep Field (1995) found a few thousand galaxies in a speck of Ursa Major. The Hubble Ultra Deep Field (2004) logged 10,144 galaxies in about 11 square arcminutes after a million seconds of exposure. The eXtreme Deep Field (2012) stacked a decade of frames on the same spot and pulled out 5,500 in a patch less than half the size. Tile any of those across the sky’s 148.5 million square arcminutes and you get a few hundred billion — not two trillion. The trillions only appear when you add the galaxies the model says you’re missing.

The challenge from past Pluto

Here is the twist that almost no answer box mentions. In January 2021, a team led by Tod Lauer and Marc Postman used NASA’s New Horizons spacecraft — the one that flew past Pluto — to test the two-trillion figure directly. From more than 6.4 billion kilometres out, where the sky is about ten times darker than anything Hubble can see, they measured the faint, combined glow of all the galaxies too distant to resolve: the cosmic optical background.

If two trillion galaxies were really out there, that glow should have been bright. It wasn’t. Lauer’s summary: take every galaxy Hubble can see, double it, and that’s all the light there is — not ten times more. That points to hundreds of billions, the older pre-2016 number, not trillions.

An honest caveat. NASA later issued a correction: the New Horizons measurement constrains the total light all galaxies emit, not the galaxy count directly. It doesn’t flatly disprove two trillion — but it leaves very little room for a vast unseen population. The dispute is real and unresolved, which is precisely why the “2 trillion” headline deserves an asterisk.

“Observable” is doing a lot of the work

Every number on this page describes the observable universe — the sphere of space whose light has had time to reach us since the Big Bang. That sphere is about 93 billion light-years across, far wider than its 13.8-billion-year age would suggest, because space itself has stretched while the light travelled.

Beyond that horizon, there is almost certainly more — possibly far more, possibly an infinite amount. We will never count it, because its light has not arrived and, due to cosmic expansion, never will. So even a perfect answer would only ever be “how many galaxies we can ever see” — not how many exist.

What Webb is changing

Since 2022, the James Webb Space Telescope has been finding early galaxies that look too big and too mature for how soon after the Big Bang they appear — some seen at redshifts beyond z ≈ 10, within a few hundred million years of the beginning. That matters here because the two-trillion figure leaned on a model of how the galaxy population behaved in exactly that early era. [VERIFY: specific record-holders shift as new JWST spectroscopy is confirmed.]

If the early universe built galaxies differently than assumed, the extrapolation that produced two trillion has to be re-run. Webb isn’t settling the count — it’s reshaping the model the count depends on. Which is the honest state of the question: the headline number is being actively rewritten right now.

Why a tidy answer keeps slipping away

Notice that the number has only ever moved in the direction of whatever we just learned to measure. Before deep fields, the guess was “a hundred billion or so.” Deep imaging pushed it up. A bolder model pushed it to two trillion. A darker vantage point past Pluto pulled it back down. Each revision wasn’t a mistake corrected — it was the limit of an instrument, made visible.

That’s the real lesson hiding inside a trivia question. “How many galaxies” isn’t a fact waiting to be looked up; it’s a running tally of how far our best telescope can presently see, dressed up as a number. The figure will keep changing not because the universe is changing, but because we are.

So when someone asks you how many galaxies there are, the most accurate reply isn’t a number — it’s “as of which telescope?” For the companion case where we genuinely can’t count the things inside a single galaxy, see how many stars are in the Milky Way.

So — what’s the real answer?

Both figures are defensible, and they answer different questions:

If you need one number for a quiz, “up to 2 trillion in the observable universe” is the conventional citation. If you want the number that survives the most direct measurement, it’s a few hundred billion. Anyone who hands you a single confident figure without that caveat is repeating a headline, not reporting a count.

Game: name that galaxy

Whatever the total, every galaxy falls into a handful of shapes Edwin Hubble sketched out in 1926 — his “tuning fork.” Spiral, elliptical, irregular. Can you call them?

Common questions

Sources

Keep exploring