How Far Away Is the Moon, and How We Know It

The Earth-Moon distance is usually quoted as 384,400 km, and that figure is both correct and not a measurement anybody took.

The earlier version of this post listed four methods for finding it and spent one sentence on each. One of the four does not work, which is worth saying plainly rather than leaving a reader to try it.

The other thing worth saying is that the technique does much more than this post credited. It watches the Moon move away from us, and it is one of the sharpest tests of gravity we have.

There Is No One Distance

Start with what the famous number is. It is the mean distance, about 384,400 km or 238,855 miles, which is roughly 30 Earth radii.

The Moon is not usually anywhere near it. At closest it comes to about 225,309 miles. At farthest it drifts out to about 251,903 miles.

So the honest range is a spread of more than 26,000 miles, and any picture showing a fixed gap is drawing a fiction.

Even the word "elliptical" is only half the explanation, and it is the half the earlier version gave.

The Moon orbits in a plane tilted about 5 degrees to the plane Earth travels in, and the Sun pulls on it continuously. Those two facts add a wobble on top of the ellipse.

The wobble is not random and it is not small. The largest piece swings the orbit by more than a degree, which moves the distance by thousands of kilometres on its own.

The consequence is that two full orbits rarely bring the Moon back to the same distance. A model is needed, not a table.

It is also getting further away

The orbit is not a closed loop. The Moon is drifting outward by about 3.8 cm every year.

That number was not calculated. It was measured, by timing light bouncing off hardware left on the surface, and it is one of the first things the method produced.

It happens because tides in the Earth's oceans, raised by the Moon and Sun, spin the Earth slightly ahead of the Moon.

The Earth drags the Moon along with it, and the energy has to come from somewhere.

It comes out of the Earth's spin, which is why the day gets a little longer as this goes on.

How a Laser Measures the Moon

Now the measurement itself, which is stranger than the summary suggests.

Five small panels sit on the Moon. Three were set down by Apollo 11, 14 and 15 between 1969 and 1973. Two rode on Soviet Lunokhod rovers.

Each panel is a slab of corner cubes. A corner cube is three mirrors meeting at a right angle, cut out of solid glass.

Light that hits one bounces straight back where it came from, whatever angle it arrived at. No power supply. No moving part. No aiming.

They are not mirrors in the usual sense. A flat mirror reflects at one angle only, so a flat mirror on the Moon would send the pulse straight past Earth except at rare moments.

A close view of the Apollo retroreflector array, a panel of corner cubes with no power supply and nothing to move. Smithsonian National Air and Space Museum and Mark Avino, CC0, via Wikimedia Commons

The Apollo 11 and 14 panels hold 100 corner cubes each, cut from fused silica and about 3.8 cm across. Apollo 15 carried 300, which is why it is the array most often hit.

The retroreflector array deployed by the Apollo 11 crew. The tilted panel is the array, placed so its face stays roughly aimed at Earth as the Moon turns. NASA, public domain, via Wikimedia Commons

Why almost every shot misses

This is the part that makes the numbers look impossible, and it is where the earlier version stopped telling the story.

Light spreads as it travels, so the pulse arrives covering an area roughly 20 km across by the time it reaches the Moon.

Most of it misses the panel entirely. The loss grows as the fourth power of distance, which is brutal at this range.

The overall round-trip loss works out to around one part in 10 to the 21st. For most pulses, nothing comes back at all.

On the older systems, a single range measurement collected fewer than 100 returning photons.

Now they fire thousands. The telescope at Apache Point gathers the returns onto a detector that can find single photons, and it reaches a statistical uncertainty of about 1 mm.

A ground laser ranging station in use today. This is the other half of the measurement: the telescope that fires the pulse and times the handful of photons that come back. H. Raab, CC BY-SA 3.0, via Wikimedia Commons

It takes five to twenty minutes of ranging to build one measurement, because the returns are averaged into a single figure.

And the ground stations have now overtaken the hardware on the Moon. Regolith shifting under the panels, and the panels expanding and contracting with the lunar day, are what limit the answer.

Forty years of station work was needed before the reflectors themselves became the bottleneck.

What Fifty Years of Timing Produced

The earlier version gave no results at all. Here they are.

The headline is gravity. Timing the Moon to a millimeter over decades lets physicists test whether mass falls at the same rate whether it is at rest or moving.

The best current limit on that test comes out at a part in 10 to the 13, which is where laboratory experiments on torsion balances and space ranging meet.

There is a second test hiding in the same data. The Moon's path would be distorted if the pull of gravity changed over time, and it has not been found to.

A map of the Moon marking the retroreflector sites, drawn from Lunar Orbiter Laser Altimeter data. Five small panels have set the Earth-Moon distance for more than half a century. NASA, public domain, via Wikimedia Commons

The reflector that got lost

One of the five tells a story worth knowing, because it shows what these panels are actually good for.

The Lunokhod 1 rover landed in 1970 carrying a French-built reflector. Some returns were recorded, but the coordinates were never published and the site was effectively lost.

For about forty years nobody could range to it. Then an image from the Lunar Reconnaissance Orbiter camera in 2010 fixed the rover's position to about 100 m.

A Lunar Reconnaissance Orbiter camera frame showing the Lunokhod 1 rover and the reflector panel on its left side. Finding the rover in this image is what made that reflector usable again after about forty years. NASA, GSFC and Arizona State University, public domain, via Wikimedia Commons

The Apache Point system picked up returns from it shortly after. It is now ranged to regularly alongside the others.

That is not a failure of the method. It is what the method is for: tying a ground measurement to a specific place on another world.

A sixth array arrived in 2025 on a commercial lander. Five more are planned, including one for the south pole.

How It Was Done Before Any Reflector

Every method above needs one thing: a length you already know, to measure against.

Around 270 BC, Aristarchus supplied one out of geometry. He timed how long the Moon took to cross the Earth's shadow during a lunar eclipse.

He knew the Moon's trip through the shadow and the Moon's full circuit took a fixed ratio of time. Two ratios, one unknown length, and the eclipse supplies it.

His answer was about 60 Earth radii. The number this article opens with is also about 60 Earth radii, which is a better result than it has any right to be.

The 1676 diagram accompanying Ole Rømer's account of timing the eclipses of a moon of Jupiter. Ranging a world by the travel time of its light predates every reflector on the Moon by three centuries. Ole Rømer, public domain, via Wikimedia Commons

Two centuries later Hipparchus used a different baseline: the Moon's own shift against the background stars between two observatories.

That shift is the same parallax a finger makes when you close one eye. Two well-separated observers see the Moon in two slightly different places, and the difference between those places is a baseline of known length.

Timing light was the next idea

In 1676 Ole Rømer noticed something about a moon of Jupiter. Its eclipses arrived earlier when Earth was on the near side of the Sun, and later when it was on the far side.

He had accidentally measured that light has a speed. The delay across the Earth's orbit came out near twenty minutes.

This is the same logic laser ranging uses. Send something out, time the return, divide. Rømer did it with a moon of Jupiter because that was what he could see.

Aristarchus used an eclipse, Hipparchus used parallax, Rømer used a clock, and the Apollo panels use a stopwatch. Same idea, four times, over two thousand years.

What astrometry is and is not

The earlier version listed astrometry as a fourth method beside the others.

It is not a method for distance. It is the measurement of where things are in the sky, and it needs a baseline just like the rest of them.

Modern astrometry using radio telescopes is part of how the Earth-Moon system is pinned down, but it is doing parallax work, not something separate.

The Method That Does Not Work

Back to the radar. It was listed third in the earlier version, between trigonometry and astrometry, as though it belonged there.

Radar from the ground reaches the Moon, and it does not come back as a usable echo at ranging wavelengths.

The reason is the surface. The Moon is covered in fine, dry grit that scatters radar energy in every direction instead of reflecting it straight back.

A diffuse return has no clean arrival time. You cannot time what you cannot isolate in time, and a millimeter range needs the arrival pinned to about a millionth of a second.

The numbers settle it. Earth-based radar needs a megawatt-class transmitter and the largest dishes in the world, and the best of it resolves range to 7.5 m.

A two-metre telescope on Earth does better than that by a factor of roughly 7500.

So radar is not a rougher or older version of the same trick. It is a different instrument doing different work.

What radar is good for

It is still one of the best tools we have for the Moon. It maps surface roughness and the scattering behavior of the soil, and it looks into craters that never see sunlight.

It can sense how deep the dust layer goes, because the signal penetrates a little way before bouncing.

None of that needs a sharp arrival time, so none of it is limited by the same problem.

The honest summary is that one method measures where the Moon is to about a millimeter, and another looks at its surface in detail. They are not competing answers to the same question.

Key Terms in Plain English

These are the words in this article that would send you to a dictionary. Each one is given here the way it is actually used above.

Term	What it means here
Retroreflector	A device that sends light straight back to where it came from.
Corner cube	A block of solid glass that sends any ray of light straight back where it came from.
Parallax	The apparent shift of a near object against a far background.
Synodic month	29.53 days from one full Moon to the next.
Perigee	Closest approach of the Moon to Earth.
Apogee	Farthest point of the Moon from Earth.
Normal point	One averaged range measurement, built from minutes of photon returns.
Ephemeris	The model that says where the Moon is on any given day.
Limb	The edge of a body as seen from far away.
Recession	The Moon moving away from Earth, here 3.8 cm a year.
Delay-Doppler imaging	Making a picture from radar timing and frequency shift.

The words in this article that would otherwise send you to a dictionary.

Key Takeaways

• 384,400 km is a mean, not a measurement. The Moon ranges from about 362,600 km to 405,400 km.
• The wobble comes from the Sun, not only from the orbit being oval. The largest piece of it is worth thousands of km.
• No two full orbits end at quite the same distance, so this needs a model rather than a lookup.
• The Moon is receding by about 3.8 cm a year, measured directly by timing light off the reflectors.
• Five corner-cube panels do the work: three from Apollo, two from Lunokhod.
• Apollo 11 and 14 carried 100 cubes each; Apollo 15 carried 300, which is why it is hit most often.
• A corner cube sends light back to its source whatever angle it arrives from. No power, no parts, no aiming.
• Most pulses return nothing. Round-trip losses run near 1 in 10 to the 21st, and the pulse covers about 20 km on arrival.
• It takes five to twenty minutes of ranging to build a single range figure.
• Precision went from decimetres in the early days to about 1 mm, and the reflectors themselves are now the limiting factor.
• Gravity comes out at a part in 10 to the 13, matched with laboratory torsion balance work.
• The Lunokhod 1 reflector was lost for about forty years and recovered from a 2010 orbital image.
• Aristarchus got about 60 Earth radii from a timed eclipse in 270 BC, which is essentially the figure in this article.
• Radar from the ground never resolved distance usefully. Its best range resolution was 7.5 m against 1 mm for laser.
• Earth-based radar earns its place by mapping roughness and shadowed craters, which need no precise arrival time.

Frequently Asked Questions

Why do people quote 384,400 km if the distance changes?

Because it is a defined average tied to the size of the orbit, and it is a useful number. It is just not a distance you would measure on a given night.

Can I bounce a laser off the Moon myself?

Not usefully. The pulse needs to land on a panel about a metre across, and the round-trip loss is around 10^-21. It has been done from Earth in the sense that returning photons have been detected, but not as a way to measure anything.

How accurate is it, really?

About a millimeter of statistical uncertainty on a single normal point. At the scale of the Earth-Moon system that is close to nothing, which is why the method can see things moving a few centimeters a year.

Why does the Moon wobble if its path is an ellipse?

Because the Sun pulls on it too. The Moon is not falling around an Earth that stands alone. It is falling around two bodies at once, and the Sun's share of that pull is what wobbles the path.

Was radar ranging to the Moon ever attempted?

Radar observations of the Moon have been made for decades and are still made. Ranging it precisely was never achieved from the ground, because the lunar surface scatters the signal rather than returning a clean echo.

Is the Moon coming back?

No. The recession is caused by tidal friction and it slows the Earth's spin, not the other way round. On any timescale that matters, the gap is widening.

Key references
• International Laser Ranging Service, lunar laser ranging
• Merkowitz 2010, tests of gravity using LLR
• Nolan et al. 2026, planetary radar at Arecibo
• NASA Space Place, how far away is the Moon
• Stern, estimating the Moon distance

Images: Wikimedia Commons, with the author and license named in each caption. Ranges, precisions and dates follow the five sources above.