The universe’s first stars probably ignited 100–250 million years after the Big Bang and died within a few million years. Their ancient light can still reach Earth, but individual stars are so distant and rare that none has been conclusively identified—although cosmic gravitational lenses may finally bring one within JWST’s view.

The universe had to make stars before it could make almost everything familiar. Carbon, oxygen, silicon and iron were not present in meaningful quantities after the Big Bang. They had to be forged inside stars and returned to space before later generations of stars, rocky planets and living chemistry could exist.

That logic requires a first generation, yet no individual member of it has been conclusively identified. Astronomers call these hypothetical pioneers Population III stars: objects born from essentially pristine hydrogen and helium, before previous stars had enriched their surroundings with heavier elements.

The often quoted interval of 100 to 250 million years after the Big Bang is a useful model-based bracket, not an observed timestamp. NASA says the first stars may have formed as early as 100 million years after the Big Bang. A galaxy observed much later has provided evidence that substantial star formation was already under way by about 250 million years. Some current simulations permit the earliest examples even sooner.

The title’s second apparent contradiction is real only at first glance. A massive first star could die after a few million years while the photons it released kept travelling for more than 13 billion years. Webb would see the star during its brief life, not as it exists now.

The date is inferred rather than observed

About 380,000 years after the Big Bang, the universe cooled enough for electrons to join atomic nuclei. Space became transparent, leaving neutral hydrogen and helium gas, dark matter and small density variations. There were no stars yet. The period before the first sustained sources of light is known as the cosmic dark ages.

Gravity slowly amplified those variations. Gas collected inside small dark-matter haloes, and molecular hydrogen allowed some of it to lose heat. Once a dense core could no longer support itself, it collapsed towards a protostar. The sequence is physically plausible, but its timing depends on assumptions about halo growth, chemistry, turbulence, radiation and how rapidly the gas fragments.

A major 2023 review of first-star formation by Ralf Klessen and Simon Glover describes a field in which increasingly detailed simulations have replaced the simplest one-halo, one-star picture. Discs can fragment into several protostars. Those objects can merge, eject one another or keep accreting. Feedback from one growing star changes what its neighbours can become.

This is why the formation window should not be treated like a historical date. The 100-million-year edge is close to NASA’s “as early as” summary. The 250-million-year figure also appears in a 2018 Nature analysis of MACS1149-JD1, whose oxygen and stellar population implied star formation at that cosmic age. That galaxy was not the first stellar nursery, and its inferred history did not reveal the universe’s first individual star. It established how early an already evolving system had to begin building stars.

Population III describes chemistry, not age alone

The Big Bang produced overwhelmingly hydrogen and helium, plus traces of lithium. Astronomers refer to every element heavier than helium as a metal. A genuine Population III star therefore begins with effectively zero metals, rather than merely being very old.

The distinction matters because metals alter how a gas cloud cools. Carbon, oxygen and dust can radiate energy efficiently, helping enriched clouds break into many small dense pieces. Pristine gas has fewer cooling routes. Early calculations consequently favoured very massive first stars, often tens or hundreds of times the Sun’s mass.

The modern picture is wider. Simulations still tend to produce a top-heavy mass distribution, meaning more massive stars relative to a modern stellar population, but fragmentation can also make lower-mass objects. A star’s age, location or low iron content cannot by itself establish Population III status. The decisive condition is that it formed before any earlier stellar generation polluted its birth material.

A few million years is the massive-star answer

Massive stars spend fuel at extraordinary rates. NASA’s detailed guide to the first stars notes that a 60-solar-mass star would shine for less than one million years. A 100-solar-mass example could reach a surface temperature near 100,000 kelvins and radiate with roughly a million times the Sun’s power.

Such objects do not fade gently. Depending on mass and internal evolution, they may explode and leave neutron stars or black holes, undergo pair-instability supernovae that destroy the whole star, or collapse directly into black holes. Their ultraviolet radiation also ionises surrounding hydrogen, while their explosions distribute the first heavy elements. One short life can change the chemistry and temperature of the next stellar nursery.

The phrase “died within a few million years” therefore applies to the massive stars expected to dominate the first generation’s light. It is not a claim about every possible metal-free star. Physics allows a Population III star below roughly 0.9 solar masses to survive until today. None has been found, which is one reason a first generation dominated by much higher masses remains plausible, but the absence is not proof that no low-mass examples formed.

A dead star’s light can still be arriving

A star’s lifetime and its light’s travel time are separate clocks. Imagine a star shining for three million years and then collapsing. Photons released during those three million years continue moving outward after the source has gone dark. If some reach Webb today, the telescope records the star as it was more than 13 billion years ago.

This is not a delayed view of a surviving object. Astronomy supplies a lookback time. The more distant the source, the earlier the chapter being observed. And because the universe expanded while the light travelled, wavelengths that began mainly in the ultraviolet have been stretched into the visible and infrared bands that Webb can detect.

Distance nevertheless leaves very few photons for a telescope. An individual first star is much fainter than the combined stellar and nebular light of a young galaxy. Its apparent position may also blend with nearby sources. Webb’s large mirror and infrared instruments make the measurement less impossible, but not routine.

The best current candidates remain systems

LAP1-B shows how close the search has come, and how careful the wording must be. In 2025, Eli Visbal, Ryan Hazlett and Greg Bryan argued in The Astrophysical Journal Letters that the highly magnified object agreed with three theoretical expectations for Population III stars. It occupied a suitably small halo, could contain a top-heavy population and had the low total stellar mass expected for a tiny primordial system.

That made LAP1-B a serious candidate, not a confirmed isolated first star. A 2026 Nature study measured its gas-phase oxygen abundance at about 0.0042 times the solar value, with substantial uncertainty. That is extraordinarily primitive, but not zero. The system may preserve material enriched by the first stars shortly after they died. It does not supply an unambiguous spectrum from one pristine star.

SpaceDaily’s earlier account of the attempt to see first-born stars described the same basic evidential problem. A missing metal line can indicate a pristine source, or it can mean the observation was not sensitive enough. A compact glow may come from one star, a cluster, ionised gas, an accreting black hole or some combination of them.

Earendel showed what a foreground cluster can do

Gravity offers a way around the shortage of photons. A massive foreground galaxy cluster bends space-time and redirects light from a more distant source. Under the right alignment, the cluster acts as a natural telescope, magnifying and sometimes multiplying the image behind it.

Earendel is the clearest proof that this can expose an individual star at extreme distance. It appears in the Sunrise Arc galaxy as seen when the universe was about one billion years old. NASA’s Webb follow-up says the star was magnified by at least 4,000 and identifies it as a massive B-type star roughly one million times as luminous as the Sun.

Earendel is not a confirmed Population III star. It lived too late, in a galaxy that had already experienced chemical evolution. Its importance here is methodological. Hubble and Webb isolated a stellar point that would have been far below their normal detection limits. Nature supplied most of the effective magnifying power.

The useful lensing events are exceptionally narrow

Ordinary cluster lensing often amplifies a background source by a factor of ten or twenty. NASA’s description of the first-star strategy explains why that is not enough. The opportunity appears close to a caustic, a line in the lens map where the idealised magnification becomes enormous.

The smooth gravitational field of the cluster supplies the broad lens. Individual stars and compact objects between galaxies add microlensing. As a distant star moves relative to this network, its brightness can rise by thousands or tens of thousands of times and later fall again. The strongest interval may last months rather than the life of an observatory.

No catalogue can simply provide the coordinates of such an event in advance. Searches repeatedly monitor many massive lensing clusters, looking for points that appear, brighten or change colour near their critical curves. The method trades one limit for another: it supplies enough flux only across a tiny area of sky and during a fortunate alignment. Lensing improves the odds; it does not guarantee a detection.

JWST still needs favourable stellar physics

A 2024 Monthly Notices modelling study examined when a lensed Population III star might cross Webb’s detection threshold. Its favourable cases involved stars above roughly 100 solar masses, high magnification, surveys of numerous strong-lensing clusters and a sufficiently abundant metal-free population at redshifts above six.

The study also found an observational bias towards evolved stars with relatively cool surfaces, below about 15,000 kelvins in its promising scenarios. A star may spend only a small fraction of its life in that swollen, cooler state, but its spectrum then places more useful flux in Webb’s bands. The easiest first star to detect need not be representative of the population.

Identification would require a reliable cluster lens model, evidence that the source is truly stellar, and spectroscopy capable of separating a pristine atmosphere or nebula from an extremely metal-poor later object. Repeated observations would have to account for microlensing variability. Claims from colours alone would remain vulnerable to several different interpretations.

What a confirmed first star would establish

A secure detection would test more than the existence of Population III stars, which is already strongly implied. Its temperature and luminosity would constrain the first-star mass distribution. Its surroundings could show how ultraviolet radiation escaped into primordial gas. Its spectrum, or the chemistry of nearby material, could clarify how quickly the first metals appeared.

Even then, “first star” would be a population label, not proof that the detected object was literally the first star anywhere in the universe. It would mean an individual star formed from effectively pristine material during the era before enrichment became widespread.

The observational proposition is therefore narrow but credible. Massive first stars could have formed, shone and died inside a few million years. Their last photons could still be arriving at Earth now. Webb can collect them only if an intervening cluster happens to place one ancient star on the thinnest and most powerful part of its gravitational lens.