What if our galaxy met another galaxy?

We live in a huge group of stars called the Milky Way. Another huge group of stars, Andromeda, is coming closer.

What if the two galaxies met?

Our Sun is one of the stars in a vast galaxy, the Milky Way. Our nearest big neighbor galaxy is Andromeda, and astronomers have known for more than a century that it is heading our way.

So what if our galaxy met Andromeda? Would stars crash together? What would happen to the Sun? This is a thought experiment about a real possibility billions of years in the future, and the scientists studying it disagree about whether it will happen at all.

Galaxy interactions are common in the universe, and our own galaxy's future may include one. Since 1912 astronomers have known Andromeda is approaching; whether it will actually merge with the Milky Way depends on its sideways motion, which is hard to measure.

This thought experiment follows the evidence: what has been measured, what successive rounds of simulations predict, and what a merger would mean for stars, the Sun and the shape of the galaxy. Each step labels fact, model and idea.

Andromeda is very, very far away. On a dark night it looks like a faint, fuzzy smudge.

It is coming toward us very fast. But space is so big that it will take billions of years to get close.

Andromeda is about 2.5 million light-years away. NASA describes it in our night sky as a faint spindle several times as wide as the full Moon. It is falling toward us at about 250,000 miles per hour, fast enough to go from Earth to the Moon in about an hour.

Even at that speed, the distance is so vast that any meeting is billions of years away. Astronomers measured Andromeda's approach using the Doppler effect: its starlight is compressed because it is moving toward us.

Andromeda (M31) lies about 2.5 million light-years away. Its approach, about 250,000 mph, has been known since 1912 from the Doppler shift of its starlight. Whether that approach ends in a collision depends on M31's tangential (sideways) motion, which is extremely difficult to measure.

In 2012, a Hubble team led by Roeland van der Marel measured M31's sideways motion by repeatedly observing regions of the galaxy over five to seven years, and found it small enough that a head-on collision seemed almost certain.

If the galaxies meet, stars would almost never bump into each other. They are very far apart.

But the stars would get new paths. Our Sun might end up in a new part of the galaxy.

Galaxies are mostly empty space. NASA says the stars in each galaxy are so far apart that they would not collide with other stars during the meeting. Instead, gravity would throw stars into new paths. Simulations show our solar system would probably be tossed farther from the galaxy's center than it is today, but NASA says Earth and the solar system would not be destroyed.

Over about two more billion years, the two galaxies could blend into one big, rounded galaxy called an elliptical galaxy. That's what the simulations show; it is a model, not something anyone has watched.

In the 2012 simulations, stars do not collide but are scattered onto new orbits around the merged center; the Sun is probably flung much farther from the core. About two billion years after the encounter, the galaxies settle into a single elliptical galaxy, its stars on randomized orbits, without the Milky Way's flat disk.

These are model outcomes. The 2025 reanalysis finds that in about half of simulated futures no merger happens within 10 billion years; instead the galaxies may settle into a long, slow orbit around each other. A 2026 study using refined Gaia measurements finds a merger in about 90 percent of its simulated futures.

Gravity pulls the two galaxies toward each other.

Even though space is stretching, galaxies that are close together can still pull each other in.

Gravity pulls the Milky Way and Andromeda toward each other, along with invisible dark matter around both of them. The universe is expanding, but NASA explains that galaxies close together are still bound by gravity, so they can meet. Long ago, when the universe was smaller, such meetings were even more common.

Whether they actually crash depends on Andromeda's sideways drift. A small sideways motion means a head-on meeting; a larger one could mean a near miss.

M31 and the Milky Way are bound by their mutual gravity and that of the dark matter surrounding them, so the cosmic expansion does not separate them. Hubble's deep fields show galaxy encounters were more frequent in the past, when the universe was smaller.

The outcome hinges on M31's transverse velocity. Earlier work assumed its most likely value; the 2025 study propagated the measurement uncertainties through thousands of simulations and added the gravity of M33, which nudges Andromeda toward us, and the Large Magellanic Cloud, which shifts the Milky Way away.

Scientists used to be almost sure the galaxies would meet. Then one study said it was only as likely as a coin landing heads. A newer study says a meeting is likely after all.

They will keep measuring to find out for sure.

In 2025, a team led by Till Sawala of the University of Helsinki took a fresh look at measurements from the Hubble and Gaia space telescopes. They included the pulls of two smaller galaxies, the Large Magellanic Cloud and Triangulum (M33). Their result: only about a 50 percent chance that the Milky Way and Andromeda merge in the next 10 billion years.

The researchers say earlier calculations weren't wrong: with the same starting assumptions, they got the same results. Including the uncertainty simply opened up more possible futures.

In 2026, a Peking University team led by Hao Wu and Huawei Zhang measured Andromeda's sideways drift again, using carefully chosen bright stars in Gaia data. They found about a 90 percent chance of a merger in the next 10 billion years, most likely in about 6.5 billion years. Even they say the question isn't settled: only better measurements of Andromeda's sideways motion can decide it.

Sawala et al. (Nature Astronomy, 2025) found that, with uncertainties in Hubble and Gaia data propagated and M33 and the LMC included, merger within 10 billion years occurs in only about half of simulated trajectories. STScI's release calls the long-term future highly uncertain and the Milky Way's fate an open question.

The team reproduced the earlier results under the earlier assumptions, so the change reflects uncertainty, not error.

Wu et al. (The Astrophysical Journal Letters, 2026) traced the scatter in Gaia proper motions to sample selection, rebuilt M31's motion from supergiant stars, and found a merger probability of about 90 percent within 10 billion years (median merger time about 6.5 billion years), with a 2σ range of roughly 65 to 100 percent. They too call the question inconclusive. The key unknown remains Andromeda's transverse velocity.