1. Phobos Looks Like an Asteroid but Moves Like a Moon Born at Mars
Phobos is a small, lumpy world about 22 kilometers (14 miles) across. Deimos, the other moon, is about 13 kilometers (8 miles) across [2]. For a long time, their shapes and the way their surfaces reflect light seemed to match asteroids in the belt between Mars and Jupiter, so many scientists thought Mars had simply caught two passing asteroids [4].
The trouble is the orbits. Both moons travel above Mars's equator in almost perfectly round paths [4]. The French research agency CNRS put the problem this way: the shape and the course of the orbits contradict the capture idea, because no one had understood how Mars could capture bodies into almost circular, equatorial orbits [3].
So the answer to the title question is "we don't know yet." Space agencies describe two live possibilities: the moons may be captured asteroids, or they may be the remains of debris thrown into orbit when a large object crashed into young Mars [1]. The rest of this page looks at each story and asks how a rock sample could help scientists choose between them.
2. Why Is It Hard for Mars to Capture an Asteroid and Leave It in a Round Orbit?
The capture story has an obvious strength: Phobos really does look like an asteroid. But looking alike is a clue, not proof [4].
The weak spot is the path. An asteroid arriving from far away and getting caught by Mars's gravity would normally end up on a tilted, stretched-out orbit. To settle onto a round orbit above the equator, it would need some way to lose a lot of its energy. Tokyo Tech researchers describe explaining the moons' present orbits by capture as extremely difficult [4].
A simple way to picture it: imagine someone wandering past a school running track, stepping onto it by chance, and then jogging exactly one perfect circle forever. It could happen, but you would want a good reason. "Difficult" is not the same as "impossible," and the space agencies still list capture as a live possibility [1]. (CNRS reported one team that argued capture is excluded by the moons' surface properties [3], but the space agencies still list both ideas as open.)
3. How Could a Giant Crash Into Mars Leave Behind Material for Moons?
In Mars's northern half lies a huge, low basin called Borealis, thought likely to have been dug by a giant impact. The impact story starts from there: the crash threw out debris that spread into a disk around Mars, and moons grew from it [4].
Computer simulations tested this. In the version Tokyo Tech described in 2017, the crashing body was about one third as wide as Mars and hit at about 6 kilometers per second, which is roughly 21,600 kilometers (13,400 miles) per hour [4]. A French-Belgian-Japanese team in 2016 also modeled a collision between Mars and a body one third its size, and found that Phobos and Deimos could grow from the debris [3]. These numbers come from simulations, not from measurements of an actual crash.
A surprising detail: the simulated disk first built a moon far bigger than Phobos. CNRS describes it as about a thousand times the size of Phobos. Mars's tides then pulled most of the new moons back down onto the planet, including that large moon, within about five million years, and Phobos and Deimos are described as what remained [3].
4. What Would a Sample From Phobos Look Like Under Each Story?
The two stories predict different ingredients, and that is what makes a sample so valuable. If Phobos came from a giant impact, it should be made of material that records the conditions on young Mars. If it is a captured asteroid, its makeup should differ from Mars [1].
The impact simulations give a more detailed forecast. About half of the material is expected to come from Mars, and the rest from the crashing body. About half of the Mars share is predicted to be mantle, the rock layer under the crust, dug up from depths of 50 to 150 kilometers (31 to 93 miles) [4].
Water is another clue. The Planetary Society explains that if the moons formed in a giant impact, the heat and energy likely stripped most of their water. If they are captured asteroids, they should still hold a substantial amount of water and organic (carbon-based) materials, like those on similar small worlds [2]. These are expected differences, not guaranteed tests, and real rocks can be messy.
There is a bonus. If the impact story is right, Phobos is a time capsule from young Mars, a world thought to have once had surface conditions more like young Earth's [1].
5. How Will Japan's MMX Spacecraft Bring Back a Piece of Phobos?
MMX stands for Martian Moons eXploration. It is led by JAXA, Japan's space agency, together with NASA, the French space agency CNES, the German Aerospace Center DLR and the European Space Agency (ESA) [1]. Its plan is to land on Phobos, collect at least 10 grams of material, and carry it home in a return capsule [1].
According to the Planetary Society, MMX launched on October 19, 2026 (October 20 in Japan) [2]. ESA's fact sheet describes the plan this way: the spacecraft reaches Mars about a year after launch, orbits around Phobos for about three years, and returns the sample in Japan's fiscal year 2031 [1]. Japanese fiscal years run from April to March, so that means sometime between April 2031 and March 2032. Dates after launch can change, so check the agencies for updates.
No spacecraft has ever landed on either Martian moon, so a sample would be a first [2].
6. Try It Yourself: Picture 10 Grams and Do the Sample Math
How much is 10 grams? About the weight of two US nickels (a nickel weighs about 5 grams). Imagine a small pinch of another world's rock that weighs about as much as those two nickels.
Now apply the impact story's forecast. If half of the material came from Mars and half of that was mantle, then in 10 grams you might find about 5 grams from Mars and about 2.5 grams of Martian mantle. This is our own back-of-the-envelope arithmetic, not a prediction from the scientists. The real sample could be mixed unevenly, and the landing site may not match the moon as a whole.
You can read the original pages in the sources below, and look for news in 2031 and 2032 about what the sample showed. Which story fits will be decided by measurements, not by which one sounds nicer.