Why most moons cannot keep a moon of their own
A submoon has to orbit its parent moon without being pulled away by the planet or spiraling inward. Tidal forces—the same kind that drive ocean tides on Earth—gradually change orbits, and for a small or close-in moon these changes happen fast enough to doom a submoon. A 2022 study found that 10-kilometer-scale submoons can only survive around moons roughly 1,000 kilometers across or larger, and only if those moons orbit far from their planet [1]. Most of the Solar System's moons fail one or both conditions.
A separate simulation study reached a similar conclusion from a different angle. By tracking the gravitational interactions of moons and submoons over long periods, it found that a submoon's stable zone extends only to about one-third of its parent moon's Hill radius—the region where the moon's gravity dominates over the planet's [5]. That is a tighter leash than the zone available to the moon itself around its planet, which the same study put at about 40% of the planet's Hill radius [5].
Which moons pass the test
A handful of known moons are large enough and far enough from their planets to plausibly host long-lived submoons: Saturn's Titan and Iapetus, Jupiter's Callisto, and Earth's Moon [1]. These are not confirmed submoons—just moons where the physics does not rule them out. The same study noted that the newly discovered exomoon candidate Kepler-1625b-I, if it exists, could in principle host a large submoon, though its stability depends on unknown parameters [1].
For Earth's Moon specifically, a 2025 modeling study found that asteroid-sized submoons (around 10^15 kilograms) could have survived, but a submoon near the previously proposed upper mass limit (about 4.6×10^17 kilograms) would have pushed the Moon roughly 30% farther from Earth than its current orbit [4]. That is a reminder that a submoon is not just a passive passenger—its gravity tugs back on its parent.
Exoplanet systems may be better hosts
The Kepler-1625 system, if its moon candidate is real, offers more room. A 2025 study found that a submoon of about 10% of the Moon's mass could survive if the hypothesized moon orbited 68% farther out than the best-fit model suggests, and that giant submoons of 1.8 Earth masses are stable when the moon's orbit is wide—more than 100 times the planet's radius [4]. The same study identified a minimum in stable configurations at intermediate submoon masses around 0.1 Earth masses, likely from a trade-off between weak tides at low masses and tight Roche limits at high masses [4].
If submoons can exist, could they be habitable? A 2022 climate-modeling study of a hypothetical moon-moon in the Kepler-1625b system began exploring that question, but the work is preliminary and depends on the exomoon candidate being real [3]. The 2025 study noted that a submoon would likely need at least 0.5 Earth masses to hold a stable atmosphere and plate tectonics, and that if formation pathways favor intermediate masses, habitable submoons could be rare [4]. These are possibilities, not established outcomes.
Where astronomers might look next
A 2024 simulation study of small satellites in the Solar System used collision and ejection probabilities to identify potential candidates for subsatellites, providing targets for future observations [2]. The same work reinforced that Mars' moons Phobos and Deimos are unlikely to have formed from a single disrupted progenitor, because they would collide within about 10,000 years—a result that holds even when additional effects like the Sun's gravity and Mars' shape are included [2]. That does not bear directly on submoons, but it shows how long-term gravitational simulations are being used to test which small-body arrangements can survive.
The 2022 stability study also discussed how submoons might be detected—through photometric transits, radial velocity, or direct imaging—and how orbital stability shapes which candidates are worth pursuing [5]. For now, the answer to the question is yes in theory, but no confirmed example exists.
Sources used in this answer
Can Moons Have Moons?
10 km-scale submoons can survive only around 1000 km-scale moons on wide orbits; Titan, Iapetus, Callisto, and Earth's Moon are potential hosts, and the Kepler-1625b-I exomoon candidate could in principle host a large submoon.
From Mars' Moons Formation to Submoons Hypothesis
Simulations of collision and ejection probabilities identify potential Solar System candidates for subsatellites, and additional dynamical effects do not change the finding that Phobos and Deimos would collide within about 10,000 years under the common-progenitor scenario.
Climate modelling of hypothetical moon-moons in the Kepler-1625b system
Climate modeling of a hypothetical moon-moon in the Kepler-1625b system provides initial results on the conditions such an object might experience.
A revision of the lifetime of submoons: tidal dynamics with the Euler-Lagrange equation
Earth could have hosted asteroid-sized submoons, a submoon near the upper mass limit would have pushed the Moon about 30% farther out, and Kepler-1625-like systems can stably host giant submoons if the moon's orbit is wide.
Orbital Stability of Exomoons and Submoons with Applications to Kepler 1625b-I
Assuming circular coplanar orbits, the stability limit is 0.40 of the planet's Hill radius for exomoons and 0.33 of the satellite's Hill radius for submoons.