Why a moon can stay attached when its planet is thrown out
A planet gets ejected when gravitational encounters with other planets or passing stars fling it out of its birth system. The key point for moons is that the ejection acts on the planet-moon pair as a unit: the planet's gravity doesn't switch off just because the star is gone. Simulations of three Jupiter-mass planets show that when one is ejected, a large share of its 100 test-particle moons survive, and moons in two-body and three-body mean-motion resonances—orbital rhythms where moons repeatedly line up—often come through with the resonance intact [5].
Close stellar encounters, a different ejection route, also preserve moons. Numerical scattering experiments find that stellar encounters keep moons more effectively than planet-planet scattering across all orbital separations, and surviving moons on tighter orbits tend to retain near-circular, low-inclination paths, while wide-orbit moons end up more dynamically stirred [2]. So the answer isn't just 'possible in principle'—the dynamics of ejection actively favor moon retention in many cases.
Distance from the planet decides whether the moon makes it
The single most important variable is the moon's orbital distance. In the ejected-gas-giant simulations, moons were likely to survive at semi-major axes beyond about 200 planetary radii (roughly 0.1 AU in that setup), and the surviving moons showed a broad spread of inclinations and eccentricities, including nearly hyperbolic and retrograde orbits [5]. In the stellar-encounter experiments, survival held for moons inside about 0.4 of the planet's Hill radius—the zone where the planet's gravity dominates over the star's—and fell off beyond that; an Io-like moon at 0.008 Hill radii survived, but retention dropped past 0.4 [2].
These two results come from different setups and different definitions of 'far,' so they shouldn't be read as a single threshold. What they agree on is the pattern: tightly bound moons ride out ejection, while loosely bound ones are more easily stripped or thrown into wild orbits.
Could we actually detect a moon around a rogue planet?
Rogue planets are faint and starless, so the practical route is gravitational microlensing—the brief brightening when a foreground object's gravity bends a background star's light. Theory work shows that a free-floating planet with a moon would produce a recognizable extra bump in the light curve, and estimates the chance of catching the moon's signature at roughly 0.002% to 0.094% per event, with the best odds (about 0.094%) for Saturn-mass planets with the moon at around 43 Saturn radii [1]. Those are small numbers, but they describe a real, calculable signal rather than a purely hypothetical one.
The stellar-encounter work goes further and applies its cross-sections to the microlensing system MOA-2011-BLG-262Lb, suggesting possible original orbits for a solar-mass host between about 1.3 and 5.9 AU, with a preference near 5.2 AU—though the authors stress the uncertainties are large and the probability distribution is flat [2]. They also note that current and upcoming microlensing and direct-imaging surveys may be able to detect Galilean-mass moons around rogue planets, which would let us test how such pairs formed [2].
What a moon around a rogue planet would mean for the planet and for life
A moon isn't just a decoration on a dark world. In the ejected-gas-giant simulations, moon-planet interactions—especially when a mean-motion resonance is present—can keep a moon's interior molten for billions of years through tidal flexing, the same churning that heats moons like Europa in our solar system; the authors connect this to the possibility of life in subsurface oceans on moons of rogue planets [5]. That is a suggestion about potential habitability, not a demonstrated finding.
There are also more speculative angles. One paper proposes that a mini black hole of about 10^11 grams in a low orbit around a frozen rogue planet could, through Hawking radiation, supply a civilization's energy needs, producing a detectable gamma-ray source that a warm planet occults every 1–2 hours [4]. Separately, a geophysical-definition paper argues that rogue planets and moons should both be classified as planets, which is a labeling proposal rather than evidence about whether such pairs exist [3]. The core answer stands on the dynamical and microlensing work: rogue planets can have moons, and the ones most likely to keep them are those with tightly bound companions.
Sources used in this answer
Microlensing due to free-floating moon-planet systems
Free-floating planet-moon systems could be detected via microlensing, with moon-induced perturbations producing characteristic light-curve features and detection efficiencies estimated at roughly 0.002%–0.094%.
Planet-moon ejections in close stellar encounters
Close stellar encounters can eject planet-moon pairs while preserving moons, especially those inside about 0.4 Hill radii, and stellar encounters retain moons better than planet-planet scattering.
Dwarf Planets and Moons are Planets: Geophysical Definition of Planet
A proposed geophysical definition classifies solar system planets, exoplanets, rogue planets, dwarf planets, and moons all as planets.
Illumination of a Planet by a Black Hole Moon as a Technological Signature
A mini black hole moon in low orbit around a frozen rogue planet could supply energy via Hawking radiation and produce a detectable gamma-ray technosignature.
Survivability of Moon Systems Around Ejected Gas Giants
Simulations of ejected gas giants find moons likely survive beyond about 200 planetary radii, with broad orbital distributions and possible tidal heating of moon interiors.