What Is the Scientific Definition of a Planet?

A planet, according to the only official definition recognized by the International Astronomical Union (IAU), is a body that orbits the Sun, has enough mass for its own gravity to pull it into a roughly spherical shape, and has “cleared the neighbourhood” around its orbit. That three-part test, adopted in 2006, remains the formal standard, but it has been controversial since the moment it passed. Critics point out that it applies only to our solar system, says nothing about the thousands of worlds discovered around other stars, and rests on a concept of orbital clearing that is surprisingly hard to pin down.

The IAU’s Three-Part Test

The 2006 resolution divided solar system bodies into three categories. A “planet” must satisfy all three criteria: it orbits the Sun, it is massive enough for self-gravity to overcome rigid-body forces and achieve hydrostatic equilibrium (a roughly round shape), and it has cleared the neighbourhood around its orbit. A “dwarf planet” meets the first two but fails the third. Everything else, aside from satellites, falls into the catch-all bin of “small solar system bodies.” Under this framework, the solar system has eight planets. Pluto, Ceres, Eris, Haumea, and Makemake are dwarf planets.

The resolution was adopted by a vote of IAU members present at the 2006 General Assembly in Prague. Only a fraction of the organization’s membership was in the room, and the debate was fierce even among those who voted. The outcome has been contested by planetary scientists ever since, not because anyone doubts that Jupiter and Earth are planets, but because the dividing line between “planet” and “dwarf planet” strikes many researchers as arbitrary or poorly defined.

How the Count Has Changed Over Centuries

The word “planet” is ancient Greek for “wanderer,” originally applied to the five bright objects that moved against the fixed stars: Mercury, Venus, Mars, Jupiter, and Saturn. By the mid-nineteenth century, the count had ballooned to thirteen. Ceres, Pallas, Juno, and Vesta were all called planets when they were first discovered in the asteroid belt, alongside Uranus and Neptune.

It is sometimes claimed that asteroids lost their planet status because they shared orbits with other objects, but a review of the historical literature shows that is not what actually happened. Asteroids were broadly recognized as a subset of “planet” for about 150 years. Their reclassification came in the 1950s, driven by two developments: new theories of planet formation no longer needed asteroids and planets to share a taxonomic category, and a growing body of research showed that asteroids were geophysically quite different from the large planets.

That precedent matters because the 2006 Pluto debate echoed many of the same arguments. In both cases, the discovery of many similar objects in the same region of space forced astronomers to decide whether the category of “planet” should stretch to accommodate them or contract to exclude them. In both cases, contraction won. The count went from thirteen to nine in the early twentieth century, then from nine to eight in 2006.

What “Clearing the Neighbourhood” Actually Means

The most contested piece of the IAU definition is the requirement that a planet must have “cleared the neighbourhood around its orbit.” The resolution itself does not specify what counts as cleared or how to measure it. That vagueness has fueled two decades of argument, but researchers have proposed quantitative ways to make the criterion precise.

One approach defines a “clearing time,” the time a body would need to gravitationally sweep its orbital zone of smaller debris. An Earth-mass body orbiting a Sun-like star at Earth’s distance could clear its zone in roughly a million years. If you require the clearing time to be shorter than the host star’s lifetime, you get a minimum mass for planet status at each orbital distance. That threshold cleanly separates the eight solar system planets from everything else: the planets all exceed it by large margins, and the dwarf planets all fall far below it.

The gap is not subtle. Earth’s clearing ability exceeds the threshold by a factor of roughly a million. Even Mercury clears comfortably. Pluto, on the other hand, falls short by several orders of magnitude. In this sense, the eight-planet solar system is not a borderline call; the dividing line sits in a wide empty space between two well-separated groups. The argument is not about whether the gap exists but about whether orbital clearing is the right criterion to care about in the first place.

The Geophysical Alternative

A vocal group of planetary scientists, particularly those who study the surfaces and interiors of worlds rather than orbital dynamics, prefer a definition rooted in what a body is rather than where it orbits. Their argument: if a body is massive enough for gravity to pull it into a sphere, it has the internal differentiation, geology, and in some cases atmospheres and weather that make it interesting in the same way Earth and Mars are interesting. Under this “geophysical” definition, Pluto, Ceres, and many large moons would count as planets.

The idea has a long intellectual history. The hydrostatic equilibrium criterion in the IAU definition already acknowledges that roundness matters. The geophysical camp simply wants to stop there and drop the orbital-clearing requirement. Their definition would produce a solar system with well over a hundred planets, including Europa, Titan, and Ganymede. Critics respond that this makes the category so broad it loses practical usefulness, though supporters counter that we do not limit the word “mountain” to a manageable number of peaks.

Where Do Planets End and Brown Dwarfs Begin

The IAU’s three-part test sets a floor for what counts as a planet but says nothing about a ceiling. At the high-mass end, planets blur into brown dwarfs, objects too massive to be planets but not quite massive enough to sustain hydrogen fusion like a star. The traditional dividing line has been pegged at about 13 Jupiter masses, the threshold above which an object can fuse deuterium in its core.

Detailed models of how deuterium burning works in massive giant planets and low-mass brown dwarfs put the mass at which half the initial deuterium is burned somewhere in the range of about 12 to 14 Jupiter masses, depending on composition and how the object formed.

⁴

But deuterium burning is a gradual process, not an on-off switch, so some researchers find this boundary unsatisfying. A more physically grounded distinction may come from formation history. Observations show that companions to stars at separations of less than a few hundred astronomical units tend to cluster into two groups: a high-mass group (stellar and heavy brown dwarf companions) and a low-mass group (planets and lighter brown dwarfs), separated by a gap known as the brown dwarf desert. The high-mass group likely formed by gravitational collapse in a disk, while the low-mass group formed by gradual accumulation of solids and gas.

⁵

The trouble is that you cannot always tell how a given object formed just by looking at it. A 10-Jupiter-mass body that formed by core accretion is physically almost identical to one that formed by disk instability. Until formation pathways can be reliably diagnosed from observation, the upper boundary of “planet” will stay blurry.

The Lower Boundary and the “Potato Radius”

At the small end of the scale, the question is: how small can a planet be? The IAU definition requires hydrostatic equilibrium, which in practice means a body massive enough for gravity to crush it into a sphere. For icy bodies like the moons of the outer solar system, the transition from lumpy “potato” shapes to roughly spherical ones happens at an average radius of about 200 to 300 kilometers.

⁶

Rocky bodies need to be somewhat larger because rock is stronger than ice and resists gravitational rounding more stubbornly. One proposed lower mass limit for a nearly spherical body sits at roughly the mass of Mimas, Saturn’s smallest round moon, which is about 0.03 Earth masses.

⁷

The transition is not sharp. Vesta, for example, is nearly round but not quite. Ceres, slightly larger, is round enough to qualify as a dwarf planet. The physics that determines roundness involves a tug-of-war between gravity pulling material inward and the rigidity of rock or ice resisting deformation.

⁸

For practical classification, the potato radius provides a reasonably clear boundary. Below it, objects are irregular lumps. Above it, they are round. Whether “round” alone is enough to earn the title of planet or dwarf planet remains the subject of the geophysical-vs-dynamical debate described earlier.

The Exoplanet Problem

Perhaps the most consequential weakness of the 2006 IAU definition is that it mentions the Sun by name. That means it technically applies only to our solar system and has nothing to say about the more than 5,000 confirmed exoplanets orbiting other stars. Researchers have pointed out that this exclusion is not just a technicality: the definition is problematic precisely because it is vague and leaves exoplanets entirely out of the picture.

⁹

The IAU does have a separate working definition for exoplanets, but it has not been formally adopted with the same authority as the 2006 resolution. That working definition suggests the minimum mass or size for an extrasolar object to be considered a planet should match the solar system standard: enough mass for self-gravity to overcome rigid-body forces and for the body to clear its orbital neighbourhood.

¹⁰

In practice, exoplanet researchers rarely worry about whether their discoveries meet the IAU definition. The objects they find are usually so massive and so close to their stars that their planetary nature is not in doubt. The harder classification problem for exoplanets is not planet-vs-not-planet but what kind of planet, and this is where new categories have emerged that have no solar system analogue.

New Categories With No Solar System Match

Our solar system has a conspicuous gap between Earth, the largest rocky planet, and Neptune, the smallest gas-rich one. Exoplanet surveys have found that this gap is not typical. Many planetary systems contain worlds with sizes between Earth and Neptune that fall into two broad groups: super-Earths, which are rocky planets larger than Earth, and mini-Neptunes, which have thick gaseous envelopes over a smaller core. These two populations appear to behave differently in their orbits: mini-Neptunes follow a pattern where those closer to their star tend to have more elongated orbits, while super-Earths may follow the opposite trend.

¹¹

The dividing line between these two types sits at a radius of roughly 1.5 times Earth’s, though the transition is not a sharp cliff. Near that boundary, it becomes genuinely ambiguous whether a planet has a thin atmosphere over a rocky surface or a deep gaseous envelope with no solid ground at all.

¹²

These categories are informal and observationally driven. No governing body has defined “super-Earth” or “mini-Neptune” with the precision the IAU brought to “planet.” But they fill a practical need: astronomers studying thousands of exoplanets need a vocabulary that captures the diversity of what they find, and the old categories of “terrestrial” and “gas giant” are far too coarse.

Free-Floating Planets Without a Star

The IAU definition requires orbiting the Sun (or, by extension, a star). But observations in young star-forming regions and open clusters have turned up objects with planetary masses drifting through space without any stellar host. These free-floating planets, sometimes called rogue planets, appear to be common. They likely originate from planetary systems where gravitational interactions between planets, or encounters with passing stars, kicked one member out.

Simulations show that when a planetary system experiences a stellar flyby, one or both planets can be ejected entirely, forming free-floating planets. Some of these ejected worlds get captured by the passing star, and some even form free-floating binary planets, pairs of planetary-mass objects orbiting each other while drifting through interstellar space.

¹³

Additional modeling work confirms that the velocities and masses of these ejected objects carry signatures of the dynamical processes that flung them out of their home systems, making free-floating planets a potential window into the architecture of planetary systems we cannot otherwise observe.

¹⁴

Under the IAU definition, these objects are not planets because they do not orbit a star. Under the geophysical definition, they are. Whether to classify them as planets, sub-brown dwarfs, or something else entirely depends on which definition you adopt, and this is one of the cases where the choice actually matters for how we talk about the galaxy’s population of worlds.

Double Planets and the Earth-Moon Question

If two bodies orbit each other while both orbiting a star, and both are massive enough to be round, are they a double planet or a planet-moon system? The IAU definition does not address this. A satellite is implicitly excluded from planet status, but the resolution does not say what makes something a satellite versus a co-orbiting partner.

One proposed criterion looks at where the center of mass (barycenter) of the pair falls. If it lies outside the larger body, the two objects might be considered co-orbiting rather than one orbiting the other. By this measure, Pluto and Charon could qualify as a double dwarf planet: their barycenter sits in empty space between them. The Earth-Moon system does not qualify because the barycenter falls well inside Earth, though it is not at Earth’s center. A more expansive proposal defines a double planet as a pair of bodies orbiting a star where both are massive enough to clear their orbit and to be spherical, and where either the barycenter lies outside the larger body or the smaller body’s primary gravitational allegiance is to the star rather than to its companion.

¹⁵

Modeling of how binary planet systems form adds another angle. Pebble accretion simulations show that when two bodies grow together in a disk while accreting small particles, the smaller partner can grow faster than the larger one, pushing the pair toward equal masses over time. This suggests that nature might produce roughly equal-mass planetary pairs more readily than we might expect, making the double-planet question more than hypothetical.

¹⁶

Detection Bias and What We Have Not Found Yet

Any discussion of planet definition is shaped, often invisibly, by what we can actually detect. Current methods for finding exoplanets are heavily biased toward large planets close to their stars, because those produce the biggest signals in transit and radial-velocity surveys. A multivariate analysis of exoplanet habitability data found that about 75% of the systems surveyed show “good star, poor planet” characteristics, meaning the host stars are suitable but the detected planets are not Earth-like. Only a tiny fraction, around 0.6% of the sample, met even relaxed criteria for habitability.

¹⁷

This bias matters for the definition debate because it means our catalog of known exoplanets is a skewed sample. We have found many hot Jupiters and many super-Earths in tight orbits, but the kinds of planets most likely to test the boundaries of any definition, small rocky worlds in wide orbits, cold gas giants far from their stars, and objects near the planet-brown dwarf boundary, remain underrepresented. As next-generation telescopes fill in the gaps, the pressure on whatever definition we use will only increase. The categories astronomers settled on in 2006 were designed for eight familiar objects orbiting one familiar star. The universe has turned out to be considerably more inventive than that.

1
Icarus. The reclassification of asteroids from planets to non-planets ↩︎
2
The Planetary Science Journal. Discovery and Classification in Astronomy: Scientific and Philosophical Challenges and the Importance of a Comprehensive and Consistent Classification System ↩︎
3
The Astronomical Journal. A Quantitative Criterion for Defining Planets ↩︎
4
The Astrophysical Journal. DEUTERIUM BURNING IN MASSIVE GIANT PLANETS AND LOW-MASS BROWN DWARFS FORMED BY CORE-NUCLEATED ACCRETION ↩︎
5
Rendiconti Lincei. Scienze Fisiche e Naturali. The discovery of exoplanets: towards understanding planet formation ↩︎
6
arXiv. The Potato Radius: a Lower Minimum Size for Dwarf Planets ↩︎
7
arXiv. Classification and Nomenclature of Planets in the Mass-Radius Plane ↩︎
8
arXiv. Size and shape of a celestial body, definition of a planet ↩︎
9
The Planetary Science Journal. Quantitative Criteria for Defining Planets ↩︎
10
New Astronomy Reviews. The IAU Working Definition of an Exoplanet ↩︎
11
PubMed. Super-earths and mini-neptunes follow different orbital period-eccentricity relations ↩︎
12
The Astrophysical Journal. From Super-Earths to Mini-Neptunes: Implications of a Surface on Atmospheric Circulation ↩︎
13
The Astrophysical Journal. Free-floating Planets, Survivor Planets, Captured Planets, and Binary Planets from Stellar Flybys ↩︎
14
arXiv. Kick Velocities and Mass Function of Free-Floating Planets from Dynamical Ejection in Hierarchical Three-Body Systems ↩︎
15
International Journal of Astronomy and Astrophysics. The Moon Meets All Requirements of the IAU Definition for “Planet” ↩︎
16
Astronomy & Astrophysics. Forming equal-mass planetary binaries via pebble accretion ↩︎
17
arXiv. Multivariate Statistical Analysis of Exoplanet Habitability: Detection Bias and Earth Analog Identification ↩︎