A rogue planet is a world that drifts through interstellar space without orbiting any star. Unlike the planets in our solar system, which follow orderly paths around the Sun, rogue planets travel the galaxy alone, unlit by a host star and largely invisible to traditional telescopes. Astronomers have confirmed individual examples and predict that billions more are lurking unseen in the Milky Way, making them one of the most abundant yet elusive classes of objects in the galaxy. Finding them requires creative detection strategies, from watching how their gravity bends starlight to imaging their faint heat glow with space telescopes.
What Makes a Planet “Rogue”
The defining feature is simple: a rogue planet is gravitationally unbound from any star. It may have formed around a star and later been kicked out, or it may have formed on its own in a collapsing cloud of gas, much the way a star forms but at a much smaller scale. Researchers draw a distinction between rogue planets and brown dwarfs, even though both can wander freely. Brown dwarfs are thought to form alongside stars through the collapse of molecular clouds, while rogue planets form inside a planetary system and are then ejected. That difference in origin matters because it shapes predictions about how many free-floating objects of each type we should expect to find. Simulations suggest rogue planets could make up roughly one to five percent of a young star cluster’s total population, while free-floating brown dwarfs formed via cloud collapse are much rarer, perhaps a quarter of a percent of the number of stars in the same cluster.
How a Planet Gets Ejected
The most widely studied formation route is gravitational scattering within a young planetary system. When two or more giant planets form close enough to interact gravitationally, their orbits can destabilize over time. Computer simulations going back decades have shown that in a system with two Jupiter-sized planets, this instability frequently results in one planet being flung out of the system entirely, while the survivor settles into a tighter, more elliptical orbit around the star.1PubMed. Dynamical Instabilities and the Formation of Extrasolar Planetary Systems The process is chaotic and can take millions of years of gravitational nudging before the final ejection event, or it can happen relatively quickly after a close encounter between two massive planets.
Close encounters with passing stars offer another ejection pathway. When a star flies near a planetary system, its gravity can yank a planet free. This mechanism is especially relevant in dense stellar nurseries, where stars are packed closely together and encounters are frequent. The resulting rogue planet sails away on a trajectory determined by the geometry of the encounter, carrying whatever orbital momentum it had when the gravitational tug won out over its bond to the host star.
A third possibility sidesteps the ejection story altogether. Some free-floating planetary-mass objects may form directly from the collapse and fragmentation of a gas cloud, the same process that creates stars, just at a much smaller mass scale. Researchers have proposed that in regions blasted by ultraviolet radiation from nearby massive stars, collapsing gas cores can be eroded down to planetary masses before they finish forming. This “photoerosion” model has been invoked to explain one of the stranger recent discoveries in the field, which we will get to shortly.
Finding an Invisible World
Rogue planets are extraordinarily difficult to detect. They emit no reflected starlight and, unless they are very young and still radiating leftover heat from formation, they produce almost no light of their own. Astronomers have developed several strategies to find them anyway.
Gravitational Microlensing
This technique exploits a prediction of general relativity: any massive object bends the light of a more distant source passing behind it. When a rogue planet drifts between Earth and a background star, the star temporarily brightens in a characteristic, smooth curve. The smaller the lensing object, the shorter and fainter the brightening event. For an Earth-mass rogue planet, the event can last less than a few hours and produce an extremely tiny angular distortion. In 2020, a Polish-led survey reported what remains one of the strongest candidates for a terrestrial-mass rogue planet, detected through an ultra-short microlensing event with one of the smallest angular Einstein radii ever measured.2The Astrophysical Journal Letters. A Terrestrial-mass Rogue Planet Candidate Detected in the Shortest-timescale Microlensing Event That detection required catching a brief spike in brightness from a single background star, which gives a sense of how painstaking the search is.
Microlensing has a fundamental limitation: each event is a one-time occurrence. You cannot go back and observe the same rogue planet again because the alignment of planet, background star, and telescope is fleeting and unrepeatable. What you get is a light curve, and from it you infer the mass of the lens and the geometry of the event. The planet itself remains invisible.
Direct Imaging in the Infrared
Young rogue planets, still glowing from the heat of their formation, can be spotted directly if a telescope is sensitive enough. The James Webb Space Telescope (JWST) is particularly well-suited for this because its infrared instruments can pick up the faint thermal glow of objects only a few times the mass of Jupiter. Predictions published before JWST’s early observing campaigns estimated that planned surveys could turn up ten to twenty giant rogue planets in moderate-density star clusters and several dozen to around a hundred in denser regions like those in the Orion Nebula.3Publications of the Astronomical Society of the Pacific. Rogue Planets and Brown Dwarfs: Predicting the Populations Free-floating Planetary Mass Objects Observable with JWST These young clusters are ideal hunting grounds because the planets are only a few million years old and still radiating enough heat to register on JWST’s detectors.
Direct imaging only works for the youngest and most massive rogue planets, though. An ancient, Earth-mass rogue planet drifting through interstellar space would be far too cold and faint for even JWST to see.
Pulsar Timing
A more exotic approach uses millisecond pulsars, rapidly spinning neutron stars that emit radio pulses with extraordinary regularity. If a free-floating object on a hyperbolic trajectory passes near a pulsar, its gravity should tug the pulsar slightly, causing tiny shifts in the arrival times of its radio pulses. The NANOGrav collaboration searched its 15-year dataset for these perturbations and found no statistically significant detections, but the non-detection itself is useful: it places upper limits on how densely packed free-floating objects of various masses can be in our local region of the galaxy.4The Astrophysical Journal. The NANOGrav 15 yr Dataset: Search for Gravitational Scattering of Pulsars by Free-floating Objects in Interstellar Space Pulsar timing is not yet sensitive enough to detect individual rogue planets routinely, but it offers a population-level constraint that complements microlensing surveys.
The False Positive Problem
Space-based microlensing searches face a nagging issue: false positives that can mimic the light-curve signature of a rogue planet. A search using data from the TESS spacecraft across several observation sectors found one short-duration event whose shape looked consistent with a low-mass free-floating planet, but the implied abundance of such planets would be oddly high compared to existing models. The researchers considered alternative explanations, including stellar flares, heartbeat binary stars, and a phenomenon called centrifugal breakout, where material flung from a rapidly rotating star produces a brief brightening. None of these alternatives fit perfectly either, raising the possibility that this event represents a new class of false positive that future space-based microlensing missions will need to contend with.5arXiv. Searching for Free-Floating Planets with TESS: Results from Sectors 61-65
This is not just a theoretical headache. As missions become more sensitive and survey wider fields, the volume of short-duration transient signals will grow. Distinguishing a genuine rogue-planet microlensing event from astrophysical noise requires either very high-cadence observations (catching enough data points on the rising and falling curve), simultaneous observations from a second telescope to confirm the geometric signature, or both. The field is actively developing these validation strategies, knowing that the science depends on them.
How Many Are Out There
Early microlensing surveys suggested enormous numbers, with some estimates implying roughly two free-floating Jupiter-mass planets for every star in the Milky Way. Those headline figures were later revised significantly downward. Updated analyses cut the estimated population of unattached Jupiter-mass planets by about half.6Nature. The hunt for rogue planets just got tougher Even with the reduced estimates, the numbers remain staggering: billions of rogue planets likely inhabit the galaxy, spanning a wide range of masses from gas giants down to rocky bodies smaller than Earth.
The mass distribution is still being mapped. JWST surveys of young clusters are sensitive to objects in the roughly one-to-fifteen Jupiter-mass range, and the predicted yields from those surveys suggest rogue planets are more common than free-floating brown dwarfs in the same clusters by a wide margin.3Publications of the Astronomical Society of the Pacific. Rogue Planets and Brown Dwarfs: Predicting the Populations Free-floating Planetary Mass Objects Observable with JWST At the low-mass end, microlensing remains the only viable detection method, and the statistics there are still thin. A clearer census will require the next generation of dedicated microlensing surveys.
What a Rogue Planet Looks Like Up Close
One of the best-characterized free-floating planetary-mass objects is PSO J318.5-22, a roughly six-Jupiter-mass world drifting through the Beta Pictoris moving group, a loose association of young stars about 24 million years old. Because it is young and still hot, PSO J318 is bright enough for detailed spectroscopic study. JWST observations have revealed a pronounced absorption feature at 10 micrometers in its spectrum, which is best explained by a high-altitude cloud layer made of tiny amorphous silicon monoxide grains, each smaller than a tenth of a micrometer. The particle size and altitude are consistent with these grains acting as cloud-seeding nuclei, condensing out of the atmosphere to form clouds.7Astronomy & Astrophysics. Evidence for SiO cloud nucleation in the rogue planet PSO J318
The atmosphere itself appears to have a roughly solar carbon-to-oxygen ratio, slightly elevated metal content compared to the Sun, and signs that its carbon isotope ratio differs from the typical interstellar medium. The object’s brightness also varies over time, which makes sense if patchy clouds rotate in and out of view. Models that describe the spectrum by blending two different atmospheric columns (essentially, a partly cloudy sky) fit the data better than a single uniform atmosphere. In other words, if you could hover near PSO J318, you would see a world with a turbulent, cloud-streaked atmosphere, glowing a dull red from its own internal heat, with no star in the sky.
Can Rogue Planets Keep Their Moons
A planet ejected from its home system does not necessarily lose everything. Simulations have explored whether moons in orbit around a gas giant can survive the violent ejection process, and the answer is surprisingly encouraging. Moons in tight orbits, like Jupiter’s Galilean satellites, have a strong chance of riding along. One study found that orbital resonances between moons (the gravitational lockstep that keeps moons like Io, Europa, and Ganymede in a synchronized dance) are frequently preserved through the ejection. In simulations testing two-body resonances, seven out of ten systems retained the resonance after ejection, and in tests of three-body Laplace resonances similar to the Galilean system, six out of ten kept the full resonance intact.8Monthly Notices of the Royal Astronomical Society. Survivability of moon systems around ejected gas giants
The survival odds depend heavily on how far the moon orbits from its planet. Close-in moons, those orbiting within about 40 percent of the planet’s Hill sphere (the region where the planet’s gravity dominates), tend to retain nearly circular, low-inclination orbits and come through ejection events relatively undisturbed. Wider-orbit moons are more likely to be stripped away or left on excited, eccentric orbits. Stellar encounters, where a passing star ejects the planet, actually tend to preserve moons more effectively across a wider range of orbital separations than planet-planet scattering events do.9arXiv. Planet-moon ejections in close stellar encounters So a rogue gas giant wandering the galaxy with a retinue of moons in orderly orbits is not science fiction. It is a physically plausible outcome of ordinary dynamical processes.
Could Anything Live on a Rogue Planet
Without a star to provide warmth, a rogue planet’s surface would be brutally cold, far below the freezing point of water within a few million years of ejection. But the surface is not the only place to look. A theoretical proposal dubbed the “Steppenwolf” scenario examines whether a rogue planet could maintain a liquid water ocean beneath insulating layers of water ice and frozen gas, heated from below by geothermal energy. The idea is that a thick enough blanket of ice, supplemented by a layer of frozen hydrogen or helium, could trap enough internal heat to keep water liquid at depth.10The Astrophysical Journal Letters. THE STEPPENWOLF: A PROPOSAL FOR A HABITABLE PLANET IN INTERSTELLAR SPACE
The scenario requires specific conditions: the planet needs to be large enough to retain significant geothermal heat (driven by radioactive decay in its interior or residual heat from formation), and the ice and gas layers need to be thick enough to act as effective thermal insulation. It is not a guaranteed outcome for every ejected rocky world. But the basic physics is not exotic. We already know that subsurface oceans exist on moons in our own solar system, like Europa and Enceladus, maintained by tidal heating rather than sunlight. A rogue planet relying on geothermal heat instead of tidal forces would be doing something analogous. Whether such an ocean could sustain life is a separate, much harder question, but the existence of the liquid water itself appears physically plausible for the right kind of planet.
Jupiter-Mass Binary Objects
One of the most surprising discoveries in the rogue-planet story came from JWST observations of the Orion Nebula Cluster, which revealed pairs of free-floating planetary-mass objects orbiting each other. These “Jupiter-mass binary objects,” or JuMBOs, have component masses ranging from less than one to about thirteen Jupiter masses, with separations of roughly 28 to 384 astronomical units. Their existence was hard to explain through the standard ejection model, because kicking two planets out of a system while keeping them gravitationally bound to each other is extremely unlikely.
A more promising explanation involves the photoerosion mechanism mentioned earlier. In this picture, a collapsing gas core in a region bombarded by intense ultraviolet radiation from a nearby massive star gets whittled down to planetary mass before it finishes forming. If the core was already fragmenting into a binary (a common step in star formation), the result would be a pair of planetary-mass objects orbiting each other, born free-floating rather than ejected from anything. Modeling shows that for reasonable gas densities, the final masses of photoeroded cores land comfortably within the observed JuMBO mass range, and the wide separations match what you would expect from a fragmenting core rather than from the much tighter binaries produced by other formation routes.11The Astrophysical Journal. Formation of Jupiter-mass Binary Objects through Photoerosion of Fragmenting Cores JuMBOs blur the line between planet and failed star, and their existence is forcing a rethink of how “planet” should be defined when formation history, not just mass, matters.
What Euclid and Future Surveys Will Add
The Euclid space telescope, launched in 2023 by the European Space Agency, is already contributing. Early release observations of the Sigma Orionis cluster demonstrated that Euclid’s sharp infrared imaging can identify substellar objects down to about four Jupiter masses in young clusters. The resulting mass function showed no evidence of a cutoff at the planetary-mass end, meaning free-floating planetary-mass objects appear to keep forming all the way down to the lowest masses Euclid can detect.12Astronomy & Astrophysics. Euclid: Early Release Observations – A glance at free-floating newborn planets in the σ Orionis cluster Those results came from a single telescope pointing in a relatively clean, low-dust region. Full survey coverage of more star-forming regions should dramatically expand the known population.
Euclid was also designed with a microlensing component in mind. A proposed exoplanet legacy science program estimated that the mission could detect the cold exoplanet mass function down to Earth mass and below, extending to orbital separations all the way out to the free-floating regime. If the early microlensing estimates of a large free-floating Jupiter population hold up, this program could detect hundreds of free-floating planets.13Monthly Notices of the Royal Astronomical Society. ExELS: an exoplanet legacy science proposal for the ESA Euclid mission – I. Cold exoplanets NASA’s Nancy Grace Roman Space Telescope, expected to launch in the mid-2020s, will carry out a dedicated microlensing survey of the galactic bulge that should be even more powerful, providing the statistical heft needed to pin down how many rogue planets of each mass actually exist. Between Roman, Euclid, and continued JWST observations of nearby clusters, the next decade should transform rogue planets from a poorly counted curiosity into a well-characterized population with known demographics, atmospheric properties, and formation histories.
Where Rogue Planets Fit in the Bigger Picture
Rogue planets occupy an awkward taxonomic space. Objects formed through ejection are clearly planets by origin, but objects formed through cloud fragmentation or photoerosion share their formation mechanism with stars and brown dwarfs, just at a lower mass. The International Astronomical Union’s working definition of a planet requires it to orbit a star, which excludes free-floating objects entirely, even those that once orbited a star and were ejected. Most researchers in the field ignore this technicality and use “rogue planet” or “free-floating planet” freely, but the tension highlights a genuine scientific question: should we define planets by how they formed, by their mass, or by where they are now?
JuMBOs make this question sharper. A pair of three-Jupiter-mass objects orbiting each other in the Orion Nebula, formed from a fragmenting gas core, share no history with the planets in our solar system. Yet their masses are squarely in the planetary range. If future surveys confirm that photoerosion routinely produces free-floating planetary-mass objects in irradiated star-forming regions, the population of “planets” that never belonged to a star could rival or exceed the population of ejected ones. That would reshape the field’s understanding of what planets are and where they come from, turning what started as a curiosity about homeless worlds into a fundamental question about how matter organizes itself at the low-mass end of gravitational collapse.