Article
Aerospace Engineering · Orbital Mechanics · ⏱ ~13 min read · Last updated: 9 July 2026

Lagrange Points L4 and L5 — The Only Two That Are Actually Stable

Every two-body system — Sun and Jupiter, Earth and Moon, Sun and Earth — has five special points where a third, tiny mass can co-orbit without drifting away under simple gravity alone. But of those five, only two, L4 and L5, are genuinely stable against small nudges. The other three are precariously balanced saddle points (equilibria that hold steady in one direction but collapse in another) that any real perturbation eventually destroys. This asymmetry explains why more than 10,000 Trojan asteroids cluster at Jupiter's L4 and L5, while nothing natural sits at L1, L2, or L3.

TL;DR: Of the five Lagrange points in a two-body system, only L4 and L5 are truly stable — the other three are unstable saddle points. The Coriolis force bends drifting objects back into closed loops around L4/L5 whenever the smaller body's mass share stays below about 0.0385, which is why Jupiter's L4/L5 host 10,000+ Trojan asteroids.

1. The Five Lagrange Points

In the restricted three-body problem, a massless third body orbits in the combined gravity field of two large masses M₁ (e.g., the Sun) and M₂ (e.g., a planet), which themselves orbit their common center of mass. Five points exist in the co-rotating frame where the third body can remain stationary relative to the two large masses.

L1

Between the two masses, on the line joining them — the classic solar-observatory point (SOHO).

L2

Beyond the smaller mass, on the same line — home to JWST, permanently in shadow relative to Earth-Sun eclipses.

L3

Opposite the smaller mass, beyond the larger mass — the fictional "counter-Earth" location.

L4 / L5

60° ahead of and behind the smaller mass, forming equilateral triangles with M₁ and M₂ — the stable pair.

2. Effective Potential in the Rotating Frame

In the frame rotating with the two large bodies, gravity combines with a centrifugal term to form an effective potential. Lagrange points are its critical points — but critical points can be maxima, minima, or saddles, and only true minima (in an appropriate sense) survive real perturbations.

Effective potential: U_eff(x, y) = −Gm₁/r₁ − Gm₂/r₂ − ½ω²(x² + y²) r₁, r₂ = distances to each mass; ω = orbital angular rate L1, L2, L3: saddle points of U_eff (unstable in the radial direction) L4, L5: local maxima of U_eff — naively this looks unstable too! → The resolution requires including velocity-dependent forces, not just the static potential

3. The Stability Criterion

Linearizing the equations of motion around L4/L5 and including the velocity-dependent Coriolis force yields a stability condition that depends only on the mass ratio between the two large bodies.

Mass ratio parameter: μ = m₂ / (m₁ + m₂) (m₂ = smaller mass) L4/L5 are linearly stable if: μ < μ_crit = ½(1 − √(23/27)) ≈ 0.0385 Sun-Jupiter: μ ≈ 0.000954 ✓ stable (Trojans exist) Sun-Earth: μ ≈ 0.000003 ✓ stable Earth-Moon: μ ≈ 0.0121 ✓ stable (below 0.0385, holds L4/L5 dust clouds) Pluto-Charon: μ ≈ 0.109 ✗ unstable — no Trojans possible there

4. How the Coriolis Force Saves L4/L5

At L4/L5 the effective potential alone is a local maximum — a ball placed there would normally roll away in every direction, which looks unstable. But any object that starts to drift picks up velocity in the rotating frame, and the moment it has velocity, the Coriolis force acts perpendicular to that velocity.

Coriolis acceleration: a_cor = −2ω × v As the test mass drifts outward from L4/L5, it gains velocity v The Coriolis term deflects it sideways, curving the trajectory into a closed loop around L4/L5 rather than letting it escape → Result: stable "tadpole" orbits that librate around L4 or L5 without ever needing propulsion to stay nearby (below μ_crit) This is exactly analogous to how a spinning top resists falling over: angular momentum + a velocity-dependent force stabilizes what would otherwise be an unstable static equilibrium

5. JavaScript Stability Check

// Check L4/L5 linear stability for any two-body mass ratio
function isL4L5Stable(m1, m2) {
  const mu = m2 / (m1 + m2); // smaller mass fraction
  const muCrit = 0.5 * (1 - Math.sqrt(23 / 27)); // ≈ 0.03852
  return { mu, muCrit, stable: mu < muCrit };
}

const systems = {
  "Sun-Jupiter": [1.989e30, 1.898e27],
  "Sun-Earth":   [1.989e30, 5.972e24],
  "Earth-Moon": [5.972e24, 7.348e22],
  "Pluto-Charon": [1.303e22, 1.586e21],
};

for (const [name, [m1, m2]] of Object.entries(systems)) {
  const { mu, stable } = isL4L5Stable(m1, m2);
  console.log(`${name}: mu=${mu.toExponential(3)} stable=${stable}`);
}
// Sun-Jupiter: stable=true, Sun-Earth: stable=true,
// Earth-Moon: stable=true, Pluto-Charon: stable=false

6. Trojans, Tadpoles, and Space Colonies

Jupiter Trojans

Over 10,000 known asteroids in "tadpole" orbits around L4 (Greek camp) and L5 (Trojan camp), some large enough that NASA's Lucy mission is visiting several of them.

Earth Trojans

2010 TK7 was the first confirmed Earth Trojan, discovered librating around Earth's L4 point — proof the effect works even for very small mass ratios.

Horseshoe orbits

Some objects (like Saturn's co-orbital moons Janus and Epimetheus) follow larger "horseshoe" paths that loop around L3, L4, and L5 without settling into a simple tadpole.

O'Neill cylinders

Proposed rotating space habitats were specifically sited at Earth-Moon L4/L5 in 1970s NASA studies because station-keeping fuel needs there are near zero.

Frequently Asked Questions

Why are L4 and L5 stable but L1, L2, L3 are not?

L1, L2, and L3 lie along the line connecting the two large bodies, where the combined gravity and centrifugal force create a saddle point — stable in one direction, unstable in another, so any perturbation eventually grows. L4 and L5 form equilateral triangles with the two large bodies, and there the Coriolis force (present only for a moving object in the rotating frame) can bend a drifting trajectory back around a stable equilibrium, provided the mass ratio between the two large bodies is below about 0.0385.

What actually lives at L4 and L5 in the Solar System?

Jupiter's L4 and L5 points host over 10,000 known Trojan asteroids, split into the "Greek camp" (L4, leading Jupiter) and the "Trojan camp" (L5, trailing). Earth, Mars, and Neptune also have small populations of Trojans at their own L4/L5 points, and Saturn's moons Tethys and Dione each have smaller trojan moons sharing their orbits.

Could humans build a colony at L4 or L5?

Gerard O'Neill's 1970s space colonization studies specifically proposed Earth-Moon L4 and L5 as ideal locations for large rotating space habitats, since a station placed there requires very little station-keeping fuel to remain in the stable region — unlike L1 or L2, which need continuous small corrective burns.

Why doesn't every planet have Trojan asteroids?
Every planet-Sun mass ratio in the Solar System is well below the 0.0385 stability threshold, so L4/L5 stability itself is not the limiting factor. What varies is how much primordial material was available near each planet's L4/L5 zone and how long-lived those populations have been against collisional grinding and slow gravitational perturbations from other planets — Jupiter's immense mass and location near the asteroid belt made it especially effective at capturing and retaining large Trojan populations.
Is the L4/L5 stability truly permanent?
Strictly, the classical result is linear stability in the idealized circular restricted three-body problem. Real Solar System bodies experience additional perturbations from other planets, solar radiation pressure, and orbital eccentricity, which can slowly destabilize individual Trojans over millions of years — numerical integrations show some Jupiter Trojans do eventually escape, but the population as a whole remains stable on timescales comparable to the age of the Solar System.
What is a tadpole orbit versus a horseshoe orbit?
A tadpole orbit librates in a small closed loop entirely around a single Lagrange point (L4 or L5) without approaching the other Lagrange points. A horseshoe orbit is a larger-amplitude libration that sweeps past L3, loops around L4, sweeps back past L3 again, and loops around L5 — named for its horseshoe shape when plotted in the rotating frame. Saturn's moons Janus and Epimetheus follow horseshoe orbits around each other's positions rather than around fixed Lagrange points.
Do JWST and SOHO sit at stable Lagrange points?
No — JWST orbits L2 and SOHO orbits L1, both of which are unstable saddle points. Both spacecraft require periodic small station-keeping burns (roughly every few weeks to months) to counteract the natural drift away from these points, using onboard thrusters and a limited fuel supply that ultimately determines each mission's operational lifetime.
🔭 Open Orbital Resonance →