In 1917, when Einstein applied his new theory of gravity to the universe as a whole, the equations would not hold still. They described a cosmos that must either expand or contract. Believing, as nearly everyone then did, that the universe was eternal and unchanging, he added a term to hold it steady. Within fifteen years observation showed the equations had been right the first time.

The evidence arrived in pieces. From 1912 at Lowell Observatory in Arizona, Vesto Slipher found that the light of most “spiral nebulae” was shifted toward the red, meaning they were receding at startling speeds. In 1923 Edwin Hubble, using the 100-inch telescope on Mount Wilson above Los Angeles, picked out a pulsating star in the Andromeda nebula and applied Henrietta Leavitt’s law to get its distance. Andromeda was far outside the Milky Way. The nebulae were galaxies, and the universe was suddenly enormous.

In 1927 a Belgian priest and physicist, Georges Lemaître, put the redshifts and distances together with Einstein’s equations and concluded that space is expanding. Hubble published the decisive data in 1929: a galaxy’s speed of recession is proportional to its distance. Since 2018 the International Astronomical Union has recommended calling it the Hubble–Lemaître law.

The usual picture is a loaf of raisin bread rising in the oven. Every raisin sees every other raisin move away, and raisins twice as far apart separate twice as fast, though no raisin moves through the dough. Click any galaxy in the model above and the pattern looks the same from each.

Three points follow. There is no center; the expansion looks the same from every galaxy. There is no edge that the universe is expanding into; distances inside it are increasing, and that is all the theory says. And the expansion applies only on the largest scales. Galaxies, solar systems, people and atoms are held together by forces far stronger than the stretch and do not grow.

Run the film backward and everything was once packed hot and dense. That is the Big Bang: not an explosion at a place, but the expansion of all of space from an early state of extreme density, 13.8 billion years ago.

A hot beginning should have left heat behind. In 1964 Arno Penzias and Robert Wilson, working with a horn antenna at Bell Labs in Holmdel, New Jersey, could not get rid of a faint microwave hiss that came equally from all directions. They had found the cosmic microwave background: light released about 380,000 years after the Big Bang, when the universe cooled enough to become transparent, now stretched by expansion to a temperature of 2.725 degrees above absolute zero.

We can see only as far as light has had time to travel. That limit, the observable universe, is a sphere centered on us about 46.5 billion light-years in radius, 93 billion across. It is larger than 13.8 billion light-years because the sources have kept receding while their light was on the way. It contains hundreds of billions of galaxies at least. What lies beyond is presumably more of the same; we have no way to look.

Gravity should slow the expansion. Two teams set out in the 1990s to measure by how much, using exploding white-dwarf stars as distance markers. In 1998 both announced the opposite: distant supernovae were fainter than expected. The expansion is speeding up. Saul Perlmutter, Brian Schmidt and Adam Riess shared the 2011 Nobel Prize.

Whatever drives the acceleration is called dark energy, a name for a measurement, not an explanation. Add dark matter, the unseen mass that holds galaxies and clusters together, and the ledger reads roughly 68 percent dark energy, 27 percent dark matter and 5 percent everything we have ever seen.

Two results keep cosmologists busy. First, the expansion rate measured from nearby stars and supernovae (about 73 km/s per megaparsec) does not match the rate inferred from the microwave background (about 67). This “Hubble tension” has survived a decade of checking, including with the Webb telescope. Second, the Dark Energy Spectroscopic Instrument in Arizona, mapping tens of millions of galaxies, reported in 2024 and 2025 hints that dark energy may be weakening over time. If confirmed, the standard picture will need revising.

Sources and further reading