The Exoplanet Count Around Binary Stars Just Jumped 50% in One Study — and What We Thought We Knew Was a Geometric Illusion

NASA said it found something new in data from its exoplanet telescope. On May 4, the agency announced that an analysis of observations from the Transiting Exoplanet Survey Satellite had turned up 27 candidate planets orbiting binary star systems — worlds that the telescope's primary detection method had systematically missed for fifteen years.

The reason: every confirmed circumbinary planet ever found shared a geometric quirk. Each one happened to orbit in nearly the same plane as its host stars, which is the only orientation that produces detectable transits. All the others — the planets whose orbits were tilted relative to their binary's plane — were invisible to that method, and there was no obvious way to find them. Until now.

Thornton et al. sidestepped the coplanarity requirement by measuring how the gravitational pull of an orbiting planet causes the binary's orbital shape to rotate over time, shifting when mutual eclipses occur between the two stars. General Relativity, stellar tides, and rotational flattening account for some of that precession in close binaries, but excess motion points to an additional perturber: a planet. NASA Science arXiv

The 27 candidates represent a roughly 50 percent increase over the 18 previously known circumbinary planets — all of which were found by transit surveys because they aligned within five degrees of their binary's orbital plane. The new candidates sit at a range of inclinations that transits could never access. They range in estimated mass from roughly 12 Earth masses to around 3,200 Earth masses. NASA Science arXiv

"We analyse TESS photometry of a sample of 1590 eclipsing binaries from the Gaia DR3 Catalogue of Eclipsing Binary Candidates," the paper states. The key to the calculation was the multi-year baseline of observations per system. "The key to calculating all of these different influences is the long, rich set of observations available from TESS," said co-author Benjamin Montet, a Scientia associate professor at UNSW Sydney. NASA Science

None of the 27 have been confirmed with ground-based radial velocity measurements, which can break the degeneracy between a low-mass planet close in and a higher-mass planet on a wider orbit — both of which would produce similar timing signatures.

The scientific stakes extend beyond adding to a list. Theoretical models of circumbinary planet formation divide on what happens during the system's infancy. One class holds that the circumbinary disk dampens inclinations and produces coplanar planets — consistent with the existing sample. Another predicts that the binary's gravitational influence stirs young planets into wider, more tilted orbits. The observed sample, built entirely from transiting worlds, was unable to distinguish between these scenarios because it was drawn from only the geometry that transits can reveal. A non-transiting population — if it exists at the scale Thornton's candidates suggest — would constitute a dataset that could finally adjudicate between the models. arXiv

"Identifying transits in binary systems clearly is challenging, but we'd like to know more about the range of planets that can form around two gravitationally bound stars," Thornton said. NASA Science

TESS was designed to find transiting exoplanets. What it has also become, increasingly, is a precision timing instrument capable of detecting gravitational perturbations that reveal planets transits cannot. The mission has confirmed 885 exoplanets to date, with more than 7,900 candidates still pending characterization. The new result suggests the timing archive holds more than anyone has yet extracted from it. NASA Science

The paper is published May 4, 2026 in Monthly Notices of the Royal Astronomical Society, with a preprint available on arXiv (2512.07934). arXiv