Are There More Grains of Sand Than Stars?

Stars win, and it is not particularly close. The observable universe contains something on the order of 10 septillion stars, while Earth holds an estimated 7.5 quintillion grains of sand. That puts stars ahead by a factor of roughly a thousand or more, depending on which estimates you trust. The comparison sounds like it should be tight, given that beaches and deserts seem endless, but the sheer number of galaxies and the stars packed inside them tip the scales decisively.

How Astronomers Count Stars They Cannot See

Nobody has counted every star. The estimate works from two main inputs: how many galaxies the observable universe contains, and how many stars a typical galaxy holds. Surveys suggest the observable universe has somewhere around two trillion galaxies, though newer deep-field observations from the James Webb Space Telescope continue to refine that count. Recent JWST data, extending to unprecedented depth in mid-infrared bands, have already revealed galaxy number counts that diverge from predictions made by earlier models of galaxy formation, hinting that the census is still shifting.¹

Each galaxy holds vastly different numbers of stars. A small dwarf galaxy might contain a few billion. The Milky Way holds an estimated 100 to 400 billion. Giant elliptical galaxies can harbor a trillion or more. The average across all galaxy types, weighted by size, gives a rough figure of around 100 billion stars per galaxy. Multiply that by two trillion galaxies and you land in the neighborhood of 10 to the 23rd power, sometimes written as 200 sextillion. That figure carries significant uncertainty in both directions, but even the conservative end is vastly larger than any reasonable sand estimate.

One subtle complication is how the ratio of big stars to small stars varies across different environments. A recent study analyzing the initial mass function for low-mass stars across both dwarf and massive galaxies found that, despite some systematic differences between measurement methods, star formation processes appear highly uniform across the universe when accounting for the chemical composition of the gas from which stars form.² That uniformity is good news for estimation: it means the same approximate rules for how many small, medium, and large stars form out of a cloud of gas apply across most of the cosmos, making the total star count a bit more trustworthy.

How Anyone Estimates Grains of Sand

The sand side of the equation is surprisingly tricky, because “sand” has no single agreed-upon boundary. Geologists define sand as any granular material between about 0.1 and 2 millimeters in diameter, regardless of composition. By that definition, sand includes quartz grains on a beach, volcanic fragments in Iceland, and coral bits in the tropics. The most-cited estimate for Earth’s total grain count comes from a calculation that considers the volume of sandy deserts, beaches, continental shelves, and shallow ocean floors, then divides by the volume of an average grain. That exercise yields roughly 7.5 quintillion grains, or about 7.5 times 10 to the 18th power.

The number is not carved in stone. It depends on assumptions about how deep the sand layer extends in each environment, how much of the ocean floor is sandy versus silty, and what grain size you pick as “average.” A slightly different average diameter or a deeper assumed desert layer can shift the final answer by an order of magnitude in either direction. Even so, the ballpark consistently falls in the range of 10 to the 18th or 10 to the 19th, well short of the star count.

What Happens if You Count All the Sediment

One reason the “which is bigger” question persists is that some versions include far more material on the Earth side. Global analyses of ocean-floor sediment estimate that the planet’s oceans hold roughly 337 million cubic kilometers of sediment. That is enough to blanket the continents in a layer about two kilometers thick. If you treated every grain-sized particle in all that mud, silt, and sand as a “grain of sand,” the count would climb enormously, perhaps into the low 10-to-the-22nd range. At that point the comparison tightens considerably.

But that move changes the rules. Most of the ocean-floor sediment is fine silt or clay, not sand-sized grains. Lumping it all together is like counting every pebble and boulder alongside the sand and calling them the same thing. The honest comparison restricts itself to actual sand-sized particles in environments we typically associate with sand: beaches, deserts, riverbeds, and the sandy portions of continental shelves. On that basis, stars still comfortably outnumber grains.

There is also a huge quantity of ancient sandstone locked in Earth’s crust that was once loose sand. A study of Cambro-Ordovician sandstone deposited along the northern margin of the ancient supercontinent Gondwana, for instance, estimated that a single vast sheet of quartz-rich sandstone covered about 15 million cubic kilometers.³ That is just one deposit from one geological era. If you could somehow disaggregate every sandstone formation on the planet back into its original grains, the total would dwarf the surface sand estimate. But those grains are cemented into rock now; they are no longer “grains of sand” in any normal sense. Still, the existence of that buried inventory shows how arbitrary the boundary is.

The Universe Is Still Adding Stars, Just More Slowly

Stars keep forming. New ones ignite inside collapsing clouds of gas and dust across billions of galaxies at this very moment. But the rate has dropped sharply. An analysis of the fossil record encoded in the stellar populations of nearly 97,000 nearby galaxies showed that the universe’s star-formation rate peaked roughly five billion years ago and has since declined by about a factor of ten.⁴ That means the universe is still adding to the star column, but at a fraction of its former pace.

Over cosmic time, the star total has only gone up. Stars die, of course, as supernovae, as white dwarfs, as neutron stars, or as black holes. But most stars, especially the small red dwarfs that make up the bulk of the population, live for tens or hundreds of billions of years. Since the universe is only about 13.8 billion years old, the vast majority of stars ever born are still burning. The cumulative total grows with each passing era, even as the birthrate slows.

Sand, by contrast, is on a slower geological clock. New sand forms through weathering of rock, erosion by rivers, and biological processes like the breakdown of coral and shells. These processes take thousands to millions of years to produce meaningful volumes. Earth’s sand supply is essentially a fixed inheritance on human timescales, refreshed only on geological ones.

Humanity Is Eating Through Sand Faster Than It Forms

While astronomers worry about the cosmic star-formation rate, geologists and resource scientists have a more immediate concern: humans are consuming sand at a staggering pace. Sand and gravel are the most extracted solid materials on Earth, used primarily in concrete, glass, and land reclamation. Global extraction runs to tens of billions of tonnes per year, and it has been accelerating with urbanization, particularly in Asia and Africa.

The problem is not that sand will literally run out, but that the kind of sand needed for construction is specific. Desert sand grains are too round and smooth, polished by wind, to bind well in concrete. The construction industry needs the angular, rough-textured grains produced by water erosion in rivers and coastal environments. Those deposits are being dredged faster than natural processes replenish them, leading to riverbed erosion, collapsing shorelines, and even the disappearance of entire islands in some parts of Southeast Asia. Research into global sand-supply networks has flagged the transition from abundant, easily accessed deposits to increasingly stressed and contested ones as a growing environmental and geopolitical challenge.⁵

None of this changes the stars-versus-sand math in any meaningful way. Even extracting billions of tonnes annually barely dents the total grain count. But it adds an ironic twist: the resource we casually use as a metaphor for uncountable abundance is, in its most useful forms, genuinely scarce.

Sand on Other Worlds

Earth is not the only place with sand. Mars has vast dune fields visible from orbit. Venus may have wind-driven surface material. But the most striking sand world in our solar system is Saturn’s moon Titan. Radar images from the Cassini mission revealed enormous linear dune fields ringing Titan’s equator, covering roughly 13 million square kilometers, about 15% of the moon’s surface. Estimates based on dune geometry put the total volume of dune sand on Titan at around 327,000 cubic kilometers, and that figure does not include sand sheets that lack visible dune structures.⁶

Titan’s sand is nothing like Earth’s. Instead of quartz grains, it appears to be made of organic compounds, likely hydrocarbons and nitriles that precipitate from the moon’s thick atmosphere and accumulate on the surface. The grains are sand-sized, roughly a few hundred micrometers across, and behave aerodynamically like sand on Earth, forming familiar dune shapes despite being made of entirely different stuff in a frigid, methane-rich environment.

If you added Titan’s sand grains to Earth’s total, it would nudge the count upward but would not fundamentally change the comparison with stars. Even combining every sandy surface in the solar system would not close the gap. The universe simply has too many galaxies, each with too many stars.

Where the Popular Confusion Comes From

The “more grains of sand than stars” idea has a long cultural history, stretching back to Archimedes, who wrote an entire treatise on estimating how many grains of sand could fill the universe. The confusion in its modern form usually traces to a misquote or a misframing. Some popular treatments phrase the comparison as “there are as many stars in the universe as grains of sand on all the beaches on Earth,” which is much closer to true if you restrict the sand count to beaches only, excluding deserts and ocean floors. Beaches alone hold perhaps 10 to the 19th grains, which overlaps with some of the lower-end star estimates. That near-overlap makes for a vivid analogy, but it obscures the fact that the best current estimates put stars comfortably ahead.

Another source of confusion is the difference between the observable universe and the entire universe. All star estimates refer to the observable universe, the sphere of space from which light has had time to reach us since the Big Bang. The actual universe may be far larger, possibly infinite. If it extends well beyond the observable horizon, the true star count could be incomprehensibly higher, making the comparison even more lopsided. Nobody uses the full universe in these estimates because we literally cannot see it, but it is worth noting that the “stars win” answer is, if anything, conservative.

Why Each Star Gets Its Own Retinue

The comparison grows even more striking when you consider what orbits all those stars. Studies of planet occurrence rates around small stars, using data from the Kepler space telescope, found roughly 0.5 planets per star with radii between half and four times Earth’s radius, just in orbits shorter than 50 days.⁷ Extrapolating to wider orbits pushes the average higher. Many astronomers now believe that planets outnumber stars, making them even more numerous than sand grains by a wider margin still.

Some of those planets have their own sand. Rocky worlds with atmospheres and weathering processes can generate granular surface material. Titan, as noted, already has dune fields larger than many Earth deserts. Mars has sand seas covering thousands of square kilometers. If planetary sand were somehow added to Earth’s tally and planetary stars to the star tally, stars would still pull ahead, because each star tends to come with multiple planets but only a fraction of those planets have sandy surfaces.

What the Numbers Actually Feel Like

Part of why this question endures is that both numbers are so large they defeat intuition. A quintillion and a septillion are both “impossibly huge,” and the human brain does not naturally distinguish between them the way it distinguishes between, say, ten and ten thousand. Research on how adults process numbers suggests that our brains handle large quantities as approximate magnitudes, with accuracy limited by the ratio between the numbers being compared rather than their absolute sizes.⁸ When both quantities are in the quintillions or higher, the ratio feels negligible even when one is a thousand times larger. Our evolved sense of “a lot” tops out well before either number starts.

A useful way to make the gap tangible: if every grain of sand on Earth represented not one star but a thousand stars, you would get closer to the actual star count. Or picture this: for every single grain of sand on every beach, in every desert, and on every ocean shelf on the planet, there are roughly a thousand stars burning somewhere in the observable universe. The sand is the rounding error, not the other way around.

How Confident Are These Estimates

Both numbers carry wide uncertainty bands, but the uncertainties do not overlap enough to reverse the answer. Star counts depend on galaxy counts, which improved with Hubble and are being further refined by JWST, and on assumptions about the mix of star sizes within galaxies. The initial mass function, which describes that mix, appears to be fairly consistent across the universe, which narrows the error bars.² Still, the total star count could plausibly be half or double the central estimate.

Sand estimates are arguably less certain. We know the surface area of deserts and beaches reasonably well, but the depth of sand in many environments is poorly mapped. Submarine sand deposits and deep desert sand seas remain under-surveyed. Even so, to close the gap with stars, you would need the sand count to be off by a factor of roughly a thousand to ten thousand. That is well beyond any plausible revision. The conclusion that stars outnumber sand grains is robust against any foreseeable improvement in either estimate.