Astronomers working with Hubble deep-field data have concluded that the observable universe holds roughly two trillion galaxies, ten times the number previously accepted. That revised count, combined with estimates placing about 100 billion stars in the Milky Way alone, pushes the total stellar population to a figure that dwarfs even the most generous calculations of sand grains on Earth. The gap between the two numbers is not a close call; it spans several orders of magnitude, and the science behind it rests on galaxy-density measurements that keep growing as telescopes improve.
Why the star-to-sand ratio keeps widening
The familiar claim that stars outnumber sand grains depends on two separate estimates, and both have shifted in the same direction. On the stellar side, a 2016 analysis by Conselice, Wilkinson, Duncan, and Mortlock found that the observable universe contains about two trillion galaxies out to a redshift of z = 8, based on deep-field data presented in an arXiv preprint. That figure replaced older tallies that placed the count closer to 200 billion. If each of those two trillion galaxies holds, on average, 100 billion stars, the total stellar population lands somewhere around 10^23 to 10^24, a number so large it resists everyday comparison.
On the sand side, the math is far less certain. The U.S. Geological Survey defines sand-sized particles as those between roughly 62.5 micrometers and 2 millimeters on the Wentworth grade scale. That definition matters because grain size determines how many particles fit inside a given volume. Estimates of total terrestrial sand, including beaches, deserts, and ocean floors, typically land between 10^19 and 10^22 grains. Even the upper bound of that range falls short of the lower bound of the star count once the new galaxy census is factored in. The hypothesis that the faint-end slope of the galaxy luminosity function, when applied uniformly across two trillion galaxies, would exceed sand-grain totals by at least three orders of magnitude holds up under available data, though both sides of the comparison carry wide uncertainty bands.
Hubble data and Milky Way mass anchor the star count
The two-trillion-galaxy figure did not emerge from a single observation. Conselice and colleagues combined Hubble deep-field images with mathematical models of galaxy number density at different cosmic epochs. Their analysis, summarized in a Nature news report, showed that most of the newly counted galaxies are small, faint systems at high redshift that earlier surveys could not detect. By stacking and reanalyzing long-exposure images, and then extrapolating to still-fainter luminosities, the team inferred a much richer population of dwarf galaxies in the early universe than previously recognized.
Those additional galaxies sit near the detection threshold of current instruments, but statistically they matter. If the universe contains many more low-mass galaxies than expected, the integrated number of stars must rise as well. The models used in the Hubble analysis assume a distribution of galaxy brightness that continues to climb toward fainter systems, though not infinitely; at some point, feedback from star formation and the heating of intergalactic gas suppresses the creation of ever-smaller galaxies. Where that cutoff lies remains an open question, and it feeds directly into the uncertainty on the total star count.
Closer to home, the Milky Way provides a critical calibration point. The European Space Agency places the Milky Way’s star count at roughly 100 billion, a figure derived from measurements of stellar mass, luminosity, and the distribution of different star types. A separate peer-reviewed meta-analysis by Licquia and Newman, published in The Astrophysical Journal, constrained the galaxy’s total stellar mass to approximately 6 x 10^10 solar masses. Converting that mass into a star count requires assumptions about the initial mass function, the statistical distribution describing how many stars form at each mass. Because most stars are smaller and dimmer than the Sun, the actual number of individual stars per unit of stellar mass tends to be higher than a simple one-to-one ratio with solar masses would suggest. That pushes the per-galaxy star count upward, not downward.
Scaling from the Milky Way to two trillion galaxies is not straightforward. Many of those newly counted galaxies are dwarf systems with far fewer stars. Others are massive ellipticals that dwarf the Milky Way. Averaging across the full population introduces real uncertainty, but even conservative estimates that assign an average of 10 billion stars per galaxy still yield a universe-wide total north of 10^22, comfortably above most sand-grain estimates. More generous averages, closer to 100 billion stars per galaxy, move the total into the 10^23–10^24 range, widening the gap further.
Where the counting breaks down on both sides
No one has counted stars individually beyond a tiny fraction of the Milky Way. Every universe-wide total relies on extrapolating from luminosity functions, which describe how many galaxies exist at each brightness level. The faint end of that function, where the smallest and dimmest galaxies live, is the least constrained by direct observation. Conselice and colleagues acknowledged the uncertainty in their error bars: plus 0.7 trillion, minus 0.6 trillion around the central estimate of 2.0 x 10^12 galaxies. Future instruments with greater sensitivity at high redshift could shift that number again, either by revealing more dwarfs or by showing that the luminosity function flattens out sooner than expected.
The sand side of the equation is even less rigorous. No global survey has measured the total volume of terrestrial sand deposits. Estimates circulating in popular science rely on rough calculations of average beach depth, desert area, and ocean-floor sediment thickness, then apply typical grain sizes to convert volume into counts. Each step involves assumptions about geography, geology, and grain-size distribution. Coastal erosion, sediment transport by rivers, and human activities such as mining and construction all redistribute sand over time, complicating any attempt to pin down a single number.
Uncertainty also arises from what is being counted. The USGS definition of sand covers a wide size range, and real-world deposits mix sand with finer silt and coarser gravel. A calculation that assumes every cubic meter of a beach is packed exclusively with mid-sized sand grains will overshoot the true grain count. Conversely, ignoring vast but thin layers of sand on the continental shelves will undershoot it. The result is that published estimates span several orders of magnitude, and none can claim the kind of observational grounding that galaxy surveys now provide for the stellar side.
What the comparison really tells us
Comparing stars to grains of sand is less about pinning down a precise ratio and more about grappling with the scale of cosmic structures. The updated galaxy census suggests that the observable universe is richer in small, faint systems than astronomers once thought, implying a correspondingly larger population of stars. Meanwhile, terrestrial sand estimates remain approximate, and even generous upper bounds struggle to keep pace with the lowest credible star counts.
As new telescopes come online, particularly those optimized for deep, wide-field imaging and infrared sensitivity, astronomers expect to refine the faint end of the galaxy luminosity function and the total number of galaxies within the observable horizon. It is conceivable that better data will trim the two-trillion figure, but it is equally plausible that still more low-mass galaxies will appear in the record. On Earth, improved remote sensing and geophysical surveys may tighten estimates of sediment volumes, yet the inherent variability of landscapes will always leave room for debate.
For now, the balance of evidence favors a universe in which stars exceed grains of sand by at least several hundred to one, and likely much more. The exact ratio may never be known, but the exercise underscores a broader point: as instruments sharpen and models improve, our picture of the cosmos grows not only clearer but also, in many ways, more crowded than we imagined.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
