Betelgeuse's 2,200-day cycle finally has a candidate object, found where models said to look

Astronomers using the SPHERE instrument on the European Southern Observatory's Very Large Telescope in Chile report the best evidence yet for a companion star to Betelgeuse, the brightest star in Orion [s1c1][s1c2]. The detection matters less as a discovery than as a successful prediction: the team pointed at a specific place at a specific time because two earlier papers said the object would be there [s1c6].

Betelgeuse is a red supergiant about 548 light-years away and the 10th brightest star in the night sky [s1c3]. Its brightness rises and falls on a period of roughly 420 days, with a longer 2,200-day modulation layered on top [s1c4]. That second, slower cycle is the one that has resisted explanation for decades, and the long-standing suspicion has been that a companion star periodically rearranges the dust Betelgeuse throws off, changing how much of its light reaches Earth [s1c5]. Until now that was an inference from a light curve. There was no object.

Two studies, published in The Astrophysical Journal in 2024 and 2025 and circulated earlier as preprints, predicted that the companion would reach maximum elongation from Betelgeuse in December 2024 [s1c6]. The same studies also concluded it would be too faint for current instruments to see, so the team led by Miguel Montargès of the Observatoire de Paris - PSL requested VLT time expecting to set upper limits rather than to make a detection [s1c7][s1c8]. They imaged the predicted region and then ran post-processing developed for exoplanet searches, a difficult job given Betelgeuse's brightness and large angular size, according to Montargès [s1c9].

What came out was a signal at 6 sigma, which Montargès calls a "Betelbuddy" [s1c10]. For comparison, astronomers at NOIRLab reported an upper limit of 1.5 sigma in 2025 [s1c11], and 5 sigma is the conventional threshold for calling something an observation [s1c12]. The new result is four times the nominal significance of the previous limit [s1d1].

The team is explicit that this is not yet a confirmed bound companion. The test is to observe half a period after the first detection, when the object should appear on the opposite side of the supergiant, followed by further tracking to pin down semi-major axis, eccentricity and inclination [s1c13][s1c14]. Those parameters are what feed into the question that actually matters: Betelgeuse will eventually explode, and Montargès says interactions between the two stars could advance or delay that supernova [s1c15]. The work is published in Astronomy and Astrophysics [s1c17].

Two things to watch. First, the far-side observation; a 2,200-day cycle is about six years, so half of it is roughly 1,100 days from the December 2024 detection, while the researchers describe the next attempt as taking place next year [s1d2][s1c13]. Second, the method. Montargès says 30% of red supergiants in the galaxy show the same kind of light variation, which makes this a possible route to finding more such binaries around evolved stars [s1c16].