An exoplanet is a planet orbiting a star other than the Sun. Nearly 5,800 are confirmed to date, revealing spectacular diversity: scorching giants, super-Earths, ocean worlds, potentially habitable planets.
An exoplanet is simply a planet that does not orbit the Sun. The word seems obvious today, but it took until 1995 to confirm that it described real objects. Before that, exoplanets were a theoretical concept backed by logic — if the Sun has planets, why not other stars? — but without direct evidence.
The detection of 51 Pegasi b in October 1995 by Michel Mayor and Didier Queloz's team at the Observatoire de Haute-Provence opened a new era. It was a 'hot Jupiter' (0.46 M_Jupiter) orbiting its star in just 4.23 days at 0.05 AU — a configuration no one had predicted, forcing astronomers to entirely rethink planet formation. (An earlier detection, planets around pulsar PSR B1257+12 in 1992 by Wolszczan and Frail, is historically considered the first confirmed detection, but around a dead star — a very special case.)
Thirty years on, the catalogue has grown from one exoplanet to nearly 5,800 confirmed (NASA Exoplanet Archive, early 2025), with more than 8,000 additional candidates awaiting validation. This statistical explosion revealed that planet formation is the rule, not the exception: nearly every star in the Milky Way hosts on average at least one planet — hundreds of billions of worlds in our galaxy alone.
The diversity exceeds everything we imagined before 1995. Alongside the familiar 'Earth analogs' and 'cold Jupiters' we find hot Jupiters (gas giants at < 0.1 AU, their atmospheres evaporating), super-Earths (1-10 M⊕, with no Solar System equivalent), mini-Neptunes (3-10 R⊕, very abundant around red dwarfs), rogue planets (no host star, ejected by gravitational instability), ocean worlds, lava-covered planets — a bestiary nature composed with more imagination than science-fiction writers.
As of January 1, 2025, the NASA Exoplanet Archive lists about 5,800 confirmed exoplanets in more than 4,300 planetary systems. A third of these systems are multiple (several planets around one star). The record: TRAPPIST-1 and Kepler-90 with 7-8 planets each — already comparable to the Solar System.
Remarkable extremes in size and density: • Mass: from sub-Earth (Kepler-37b, 0.27 M⊕) to super-Jupiter (HD 100546 b, 20 M_Jup, low-mass end of brown dwarfs). • Radius: from 0.4 R⊕ (sub-Mercury) to 2 R_Jup (planets 'inflated' by extreme stellar heating). • Orbital period: from a few hours (KELT-9 b, 1.48 days) to several centuries (HR 8799 b, ~465 years). • Equilibrium temperature: from 50 K (cold Jupiters) to 4,600 K (KELT-9 b, hotter than most stars).
Crucial statistic: the Kepler and TESS missions showed that the most common planets are not Earth-like but sub-Neptunes (1.7-3 R⊕). A curious gap — the 'radius valley' — appears near 1.6-2 R⊕, probably linked to photochemical evaporation of atmospheres by stellar radiation.
The nearest exoplanets. Proxima Centauri b, at 4.24 ly, discovered in 2016 by radial velocity: a minimum 1.17 M⊕ super-Earth, in the habitable zone of its modest red-dwarf star. Barnard b at 5.96 ly, Ross 128 b at 11 ly. The closest potentially habitable exoplanet remains Proxima b, and no mission is feasible for decades — the Voyager probes would take 75,000 years to reach it.
Classification works by detection method and by physical nature.
By detection method:
• Transits (~75% of discoveries). When the planet passes in front of its star, it blocks a fraction of the light, observable as a periodic dimming. Flagship missions: Kepler (NASA, 2009-2018, 2,700+ confirmed), TESS (NASA, 2018-present, 400+ confirmed), CHEOPS (ESA, 2019-present, characterization), upcoming PLATO (ESA, launch 2026, search for Earth analogs).
• Radial velocities (~20%). Measures the Doppler wobble of the star under planetary gravity. First method to work (51 Peg b). Emblematic spectrographs: ELODIE (historic), HARPS (La Silla, 3,300+ confirmed), ESPRESSO (VLT, m/s precision), upcoming HIRES (ELT, 2029+, targeting 0.1 m/s).
• Direct imaging (~2%). The planet is photographed directly while the star's light is blocked (coronagraph). Works only for young, massive planets far from their star. Instruments: SPHERE (VLT), GPI (Gemini), JWST NIRCam. Landmark images: HR 8799 (4 planets directly seen).
• Gravitational microlensing (~3%). Relativistic effect amplifying the light of a background star when the planet passes in front. Very sensitive to distant planets. Roman Space Telescope (NASA, launch 2026) should find thousands.
By physical nature:
• Terrestrial / rocky (R < 1.6 R⊕). Density > 3 g/cm³. • Super-Earths (1-10 M⊕). Rocky or water-rich variants. • Mini-Neptunes / sub-Neptunes (2-4 R⊕). Significant hydrogen envelope. • Gas giants (Jupiters, Saturns, hot Jupiters, warm Jupiters). • Rogue planets. Without a host star, ejected or formed directly by cloud collapse.
One subfamily has captured the public imagination: compact multi-planet systems like TRAPPIST-1 (7 planets, all in or near the habitable zone), detected in 2017, the ideal candidate for atmospheric biosignature searches with JWST.
Detecting an exoplanet remains a monumental challenge: even a Jupiter at 1 AU around a Sun-like star is about 10⁹ times less luminous than the star itself, and at 30 light-years the 0.1 arcsecond angular separation pushes current optics to their limit. We therefore work around this contrast with indirect methods.
Major space missions. Kepler (NASA, 2009-2018) continuously observed 150,000 stars in Cygnus and revolutionized exoplanet science by statistically proving planet abundance. TESS (NASA, 2018-present) maps the whole sky for exoplanets around nearby bright stars, the best candidates for follow-up characterization. CHEOPS (ESA, 2019-) refines radius measurements of already-detected planets. JWST (NASA/ESA/CSA, 2022-) characterizes exoplanet atmospheres with infrared spectroscopy and has already detected H₂O, CO₂, SO₂ and silicate clouds in several hot-Jupiter and sub-Neptune atmospheres like K2-18 b.
Upcoming missions. PLATO (ESA, 2026): systematic search for Earth analogs in the habitable zone of solar-type stars. Ariel (ESA, 2029): atmospheric spectroscopy of 1,000 exoplanets. Roman Space Telescope (NASA, 2026-2027): wide-field microlensing and coronagraph to directly image cold giants.
Atmospheric characterization. The crucial step that transforms a 'bright dot' into a 'world'. We analyze starlight filtered through a planet's atmosphere during transit (transmission spectroscopy), revealing molecular absorptions: H₂O, CH₄, CO₂, O₂, O₃. JWST has detected exceptional signatures, and WASP-39 b became in 2022 the first exoplanet with an unambiguous CO₂ detection.
And for the amateur? A few exoplanets are technically accessible to amateur photometry (transit observation) with a 200 mm+ telescope and a CCD/CMOS camera, for bright-star transits like HD 189733 b or WASP-43 b. Citizen programs like ExoClock centralize these observations. Otherwise, you can simply spot the host stars in the sky with our sky map — 51 Pegasi, Tau Ceti, TRAPPIST-1, Proxima Centauri are all stars you can point at.
Several objects come conceptually close but remain distinct.
Brown dwarf. An intermediate body between a giant planet and a star (13-80 M_Jup), which does not fuse hydrogen but can fuse deuterium. The lower 13 M_Jup limit is set by the IAU to distinguish brown dwarfs from giant exoplanets. Brown dwarfs can orbit stars or drift freely.
Rogue planet. A planetary-mass body not bound to a star — formed as a planet then ejected, or formed directly by cloud collapse. Technically a 'free-floating planetary-mass object'. Recurring debate: should they be called 'exoplanets'? Convention: yes if ejected from a system, no if formed in isolation.
Dwarf planet. A Solar System object, spherical, orbiting the Sun, not having cleared its zone (Pluto, Ceres, Eris). By definition, not an exoplanet. But the question 'do exo-dwarf planets exist?' is open: none have been detected — they would be too faint.
Protoplanetary / transition disk. Not a planet but a structure in formation from which planets emerge. Planets forming in disks (proto-planets) have occasionally been directly imaged, like PDS 70 b and c (2018-2019) with VLT.
Exomoon. Satellite of an exoplanet. No indisputable confirmation to date (April 2026), but several debated candidates (Kepler-1625 b-i, Kepler-1708 b-i).
Proxima Centauri b, just 4.24 light-years away — around the nearest star to the Sun, a red dwarf. Discovered in 2016 by Guillem Anglada-Escudé's team with HARPS, it has a minimum mass of 1.17 M⊕ and orbits at 0.0485 AU in 11.2 days, within the habitable zone of its modest star. Its radiation environment is problematic (Proxima emits violent flares that may have sterilized the surface), but it is currently the top target for direct-imaging projects with future ELTs and LUVEX. The Voyager probes would take about 75,000 years to reach it at their current speed.
About 5,800 confirmed exoplanets at the start of 2025, according to the NASA Exoplanet Archive, in more than 4,300 planetary systems. Plus more than 8,000 candidates awaiting validation, mainly from TESS and Kepler. The figure grows by several hundred per year, and with PLATO (2026), Roman (2026-2027) and Ariel (2029) coming online, the catalogue should explode: tens of thousands more are expected in the next decade. Statistically, the Milky Way likely contains several hundred billion exoplanets — more than its stars.
Potentially, yes — but 'habitable' is not synonymous with 'inhabited'. The 'habitable zone' concept defines the orbital range where liquid water can exist on the surface of a rocky planet at reasonable atmospheric pressure. About forty confirmed exoplanets sit in their star's habitable zone (TRAPPIST-1 e, f, g, Kepler-452 b, Proxima b...). But actual habitability also requires a suitable atmosphere, a protective magnetic field, active geology, and chemistry compatible with life as we know it. No biosignature has been confirmed to date. JWST and future telescopes (HWO, LUVEX) target exactly this question.
That's an observational bias — not a reflection of reality. Detection methods are more sensitive to massive planets close to their star: a hot Jupiter at 0.05 AU induces a deep transit and huge radial velocities, easily detected. An Earth at 1 AU causes a 0.008% transit and 0.1 m/s velocity shifts, at the limit of the best current spectrographs. Corrected statistics show hot Jupiters are actually fairly rare (< 1% of stars host one), whereas super-Earths and sub-Neptunes are the most abundant population (30-50% of stars). PLATO and HIRES will finally give us access to the 'Earth-like' population around Sun-like stars.