Margo Thornton was staring at the orbits of 1,590 pairs of stars when she noticed something that the physics couldn’t explain. In system after system, the elliptical dance of two stars around each other was shifting faster than general relativity, tidal forces, and stellar spin could account for. Something unseen was pulling on them.
Thornton, a researcher at the University of New South Wales, and her colleagues believe they know what that something is: planets. Twenty-seven of them, each orbiting a pair of stars at once. Their findings, published in Monthly Notices of the Royal Astronomical Society in May 2026, represent the largest single batch of circumbinary planet candidates ever reported, and they were found using a detection method no team had previously deployed at this scale.
If even a fraction of these candidates hold up to follow-up observation, the discovery would reshape what astronomers know about where planets can form and survive.
Why circumbinary planets are so hard to find
A circumbinary planet orbits two stars instead of one. The concept is familiar from science fiction (the twin sunsets of Tatooine in Star Wars), and the UNSW team timed their publication to May the 4th, the unofficial Star Wars holiday, as a deliberate nod. But in reality, these worlds are extraordinarily rare in the astronomical catalog. Before this study, only about 14 to 15 circumbinary planets had been confirmed, nearly all of them discovered by NASA’s now-retired Kepler space telescope using the transit method, which watches for a planet crossing in front of its host stars.
The problem is that transits in binary systems are messy. Two stars create complicated, overlapping light patterns, and a planet’s orbit around a binary pair is gravitationally irregular, making repeat transits hard to predict and easy to miss. That difficulty is a major reason the confirmed count has stayed so low for over a decade.
A different way to look
The UNSW team tried something different. Instead of hunting for transits, they tracked a phenomenon called apsidal precession: the slow rotation of a binary star’s elliptical orbit over time. Think of it as the long axis of the oval gradually spinning like the hand of a clock.
Physicists can predict how fast that rotation should happen based on three factors: general relativity, tidal interactions between the two stars, and each star’s spin. Thornton’s team calculated the expected precession rate for each of the 1,590 eclipsing binary systems in their sample, then compared it to what NASA’s Transiting Exoplanet Survey Satellite (TESS) actually measured.
In 27 cases, the observed precession was significantly faster than predicted. The team argues that the excess is best explained by the gravitational tug of an unseen companion, most likely a planet, orbiting the binary pair from farther out. Six additional systems showed even larger excesses consistent with brown dwarfs or low-mass stars rather than planets.
“No one had tried this approach at scale before,” senior author Ben Montet said in a UNSW media release distributed through the AAAS platform. Thornton described the excess precession as evidence of “an unseen body tugging on the binary.” The idea that apsidal precession could betray the presence of a hidden planet had been discussed in theoretical work before this study, but the UNSW team is the first to systematically apply it across a large sample of binary systems.
The team’s target list was drawn from the Gaia Data Release 3 eclipsing-binary candidate catalog, then cross-referenced with TESS light curves. An initial preprint appeared in December 2025, with a revised version posted in March 2026 before formal journal acceptance.
Candidates, not confirmations
The word “candidates” matters. None of the 27 systems have been independently confirmed as hosting a planet. Apsidal precession infers a hidden companion from orbital mechanics; it does not directly image the planet or catch it crossing in front of a star. Alternative explanations, including unmodeled effects inside the stars themselves or systematic errors in the TESS light curves, have not been fully ruled out.
Several transparency gaps add to the uncertainty. The team has not released a publicly accessible supplementary table with individual identification numbers and precession measurements for each candidate, which would allow outside groups to reproduce the analysis system by system. The custom fitting code used to model precession rates also remains unavailable. Until those tools and data are shared, independent replication will depend on the methodology description in the paper rather than direct verification.
The six higher-mass companions sit in an additional gray zone. Their minimum masses land near the boundary between giant planets and brown dwarfs, and without tighter constraints, their true nature stays ambiguous.
Press coverage has generally handled the distinction well. The Guardian, among other outlets, described the findings as “potential” planets rather than confirmed discoveries, language that matches the paper itself.
What gives the method credibility
The MNRAS paper passed peer review, which means the statistical methodology and sample selection survived scrutiny by independent referees. That is not a guarantee every candidate will hold up, but it establishes a baseline of rigor.
The method also has a precedent. Earlier work on the TESS circumbinary system TIC 172900988 demonstrated that apsidal motion can be a measurable signature in systems where a planet orbits a binary pair. That prior detection combined TESS photometry with archival eclipse data, showing the underlying physics is sound even if applying it to 1,590 systems at once introduces new statistical complexity.
Both the Gaia DR3 catalog and the TESS mission data that feed the analysis are publicly funded, well-documented datasets with established quality controls. The raw ingredients, in other words, are solid. The question is whether the recipe the UNSW team built on top of them will prove reliable at scale.
Whether confirmation or false alarm, the answer reshapes circumbinary science
Confirmation will likely require radial velocity measurements from ground-based observatories or precise transit-timing observations from facilities like the James Webb Space Telescope. Those campaigns take time; individual systems may need months or years of monitoring before a definitive verdict arrives.
The stakes are significant. If a substantial fraction of the 27 candidates survive follow-up, the result would more than double the known population of circumbinary planets and establish apsidal precession as a practical new tool in the planet-hunting toolkit. It would also suggest that TESS data, now years old, still contain major discoveries waiting to be extracted with the right analytical approach.
If most turn out to be false positives driven by stellar physics rather than hidden planets, the method will need refinement. Either outcome will tell astronomers something important about how planets form and survive in the gravitational chaos around binary stars, a question that, until now, has been limited by a sample size small enough to count on two hands.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
