Bryan Konietzko (you know, one of the guys being Avatar The Last Airbender and The Legend of Korra), is working on a new project called Threadworlds, which takes place in a solar system that has 5 planets sharing the same orbit. That got me thinking, how many planets can fit into the same orbit without there being negative side effects on the planets themselves? How would their positions relative to one another affect this?

  • 1
    $\begingroup$ Klemperer rosette. Note that all such solutions are unstable over geological timespans; a certain amount of station-keeping magic (or engine thrust) will be needed if the system is to remain stable over long periods of time. $\endgroup$
    – AlexP
    Feb 3, 2018 at 0:04
  • 1
    $\begingroup$ On astronomical timescales, no more than two, and that two could be on a special Horseshoe orbit. Several planets may occupy each other's Lagrangian points, but those orbits are inherently unstable. $\endgroup$
    – Alexander
    Feb 3, 2018 at 0:06
  • 9
    $\begingroup$ Actually having a shared orbit with another planet would disqualify planets from being defined as a planet. $\endgroup$ Feb 3, 2018 at 0:11
  • 3
    $\begingroup$ @A.C.A.C. If you grew up on a world where there were multiple objects in the same orbit as yours, you would call yourself a planet, and adjust the definition accordingly. We singly orbital Earthlings are prejudiced (planetist?) against planets that can't clear their own orbit. $\endgroup$
    – kingledion
    Feb 3, 2018 at 2:38
  • 1
    $\begingroup$ @A. C. A. C.: Only if you agree with the bunch of (expletives deleted) who came up with the current nonsensical "orbit clearing" criterion for planet. $\endgroup$
    – jamesqf
    Feb 3, 2018 at 4:18

1 Answer 1


Well, 5 is right out without artificial stabilization.

6 planets of equal mass (or a two sets of three planets, each with roughly equal mass) arranged in a hexagon are more stable than most configurations, but the system will still fall apart under slight perturbations. So, you'd need artificial stabilization to keep that going for geologically significant time periods.

So, if you're going to have to use artificial stabilization anyway, you can actually get pretty high numbers. In fact, just about any number of equally spaced, equal-mass planets will produce a static equilibrium up until they start actually overlapping each other. So, you could theoretically fit tens of thousands of Earths in the same orbit, if you had a sufficient system of light sails, gravity tractors and so forth to dampen perturbations.

If you're looking for something that could conceivably form naturally, or at least remain stable long-term without maintenance if it were constructed, then the practical maximum is three, where one of them is a gas giant, and two terrestrial worlds occupy the leading and trailing Trojan points. The upside of that, though, is that you could add several more Earthlike worlds as moons of the gas giant.

If you don't want to use a gas giant "anchor", then the practical maximum is 2, but there are several ways to manage it. One method is to put them in a so-called "horseshoe orbit", famously exemplified by the moons Janus and Epimetheus in our own solar system. In this arrangement, one planet is actually on a slightly smaller orbit than the other, such that it eventually laps its partner and approaches from behind. The interaction of the two worlds with each other then cause the inner planet to be pulled up to a higher orbit, and the outer planet to be pulled down to a lower orbit, so that they switch places and the cycle repeats.

Another option involves eccentricity exchange. You start with one planet on a nearly circular orbit, and the other planet on a highly elliptical orbit with the same semi-major axis. The two planets will never actually come particularly near to each other, but over a very large number of orbits the planet with the originally-circular orbit will become steadily more elliptical, while the other planet's orbit will become steadily more circular, until they have completely swapped places, and as before the cycle then repeats.

You might be able to multiply any of those numbers by two, replacing isolated worlds with double planets, as long as the inter-system separations are large enough, such that each double planet can be reasonably approximated as a point mass.

  • 2
    $\begingroup$ Do you have any links to support your claims? Not saying you're wrong, only that it helps to see where you're getting your ideas from. $\endgroup$ Feb 3, 2018 at 16:10
  • $\begingroup$ I would think a horseshoe would not be stable against outside nudges. $\endgroup$ Oct 22, 2023 at 3:32

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .