Roughly 99 percent of all species that have ever lived on Earth are extinct. That number has been repeated so often in textbooks and documentaries that it sounds like settled fact, and in broad strokes it is. But it rests on a surprisingly shaky foundation of estimates, extrapolations, and fossil-record gaps that make the true figure somewhere between “almost certainly above 95 percent” and “possibly higher than 99.9 percent.” The uncertainty reveals as much about the history of life as the number itself does.
Where the 99 Percent Figure Comes From
The estimate is not the product of one landmark study. Instead, it emerged from decades of paleontological work comparing the number of species alive today against the total number that must have existed over the roughly 3.8 billion years life has occupied this planet. The logic is straightforward: if you can estimate how many species exist at any given time, how long a typical species lasts before it goes extinct, and how long life has been around, you can work backward to a rough cumulative total. When researchers do this, the number of species that must have come and gone dwarfs the number alive now by at least two orders of magnitude.
A commonly cited total for species that have ever existed falls somewhere between 5 billion and over 50 billion, depending on assumptions about microbial diversity, soft-bodied organisms, and deep-time turnover rates. With current estimates placing living eukaryotic species at around 8.7 million, the math gives you that familiar 99-plus percent, even under conservative assumptions.1PLOS Biology. How Many Species Are There on Earth and in the Ocean? But each piece of that calculation carries enormous uncertainty, which is why the figure should be understood as an order-of-magnitude statement rather than a precise measurement.
How Long Species Tend to Last
One of the key inputs to any extinction percentage is species longevity: how long does a typical species persist before it disappears? The answer varies enormously depending on the group. Marine invertebrate species in the fossil record tend to last somewhere around 5 to 10 million years on average, though some lineages are far more durable and others far shorter-lived. Among North American fossil mammals, the numbers are tighter: hoofed herbivores like artiodactyls averaged around 4 to 4.4 million years, odd-toed herbivores like horses and rhinos lasted about 3.1 to 3.3 million years, and carnivorous mammals had the shortest runs at roughly 2.6 to 3 million years.2PubMed. Species longevity in North American fossil mammals
These lifespans matter because they establish the background tempo of species turnover. Even without catastrophic events, species are constantly appearing and disappearing. The background extinction rate for marine species over the last 540 million years or so works out to a mean of about 0.25 species going extinct per million years, though that average blends long stretches of near-zero loss with occasional spikes of much higher loss.3PubMed. A kill curve for Phanerozoic marine species Separate estimates of background extinction have landed near 0.1 extinctions per million species per year, which gives a useful baseline for judging whether modern rates are abnormal.4PubMed. Estimating the normal background rate of species extinction
The Five Great Die-Offs
Background extinction accounts for most of the total species lost over Earth’s history, simply because it operates continuously across billions of years. But the events that capture the imagination are the mass extinctions: sudden spikes where large fractions of life vanished in geologically brief intervals. Five stand out in the fossil record of complex animal life, and they have collectively reshaped Earth’s biodiversity in ways that are still visible today.
The largest was the end-Permian extinction, roughly 252 million years ago, sometimes called the Great Dying. It is commonly described as wiping out 90 to 96 percent of all marine species, but those numbers deserve scrutiny. A reanalysis of the data showed that the figure for the terminal Permian crisis alone was closer to 81 percent of marine species, with the higher estimates having been inflated by incorrectly combining data from two separate extinction pulses in the Middle and Late Permian.5Europe PMC / PNAS. Estimates of the magnitudes of major marine mass extinctions in earth history That is still catastrophic by any standard, but it illustrates how extinction statistics can drift upward through the literature once a dramatic number gains traction.
The end-Cretaceous extinction, 66 million years ago, killed off the non-avian dinosaurs along with roughly three-quarters of species. The other three major events, in the Late Ordovician, Late Devonian, and end-Triassic, each eliminated somewhere between half and three-quarters of marine species. Together, these five events account for a disproportionate share of cumulative species loss, even though they represent only brief punctuations in a timeline dominated by slower, quieter disappearances.
What the Fossil Record Misses
The 99 percent estimate is built overwhelmingly on the fossil record of hard-bodied marine animals: shells, bones, teeth, and exoskeletons. These are the organisms that fossilize readily, accumulate in sedimentary rock, and can be counted across geological time. But they represent a deeply biased sample of life’s actual diversity.
Soft-bodied organisms, the ones without shells or skeletons, rarely leave fossils at all. Exceptional preservation sites like the Burgess Shale give us occasional windows into this hidden majority, and the picture is striking: without those rare localities, roughly 80 percent of Cambrian organisms would be completely unknown to science.6Geology. Soft tissue fossil clues could help search for ancient life on Earth and other planets The Cambrian is better represented than most periods because of these exceptional sites. For most of geological time, the soft-bodied majority of life left no trace at all. Every worm, jellyfish, fungus, and microbe that lived and died without mineralized parts is essentially invisible to paleontology.
Insects pose a similar problem. They are the most species-rich group of animals alive today, with estimates in the millions of species, yet their fossil record is comparatively sparse because their small, delicate bodies rarely survive long enough to fossilize. And then there are microorganisms: bacteria, archaea, protists, and fungi, whose combined species numbers are genuinely unknown. Some estimates of total microbial diversity run into the billions of species. If even a fraction of those estimates hold, the total number of species that have ever existed balloons far beyond anything the fossil record can capture, and the extinction percentage creeps closer to 99.9 percent or higher.
The Problem of Counting Living Species
The denominator matters just as much as the numerator. If you want to know what fraction of all species are extinct, you need to know how many are alive right now, and even that number is uncertain. The most widely cited estimate puts the total at around 8.7 million eukaryotic species, with a standard error of about 1.3 million.1PLOS Biology. How Many Species Are There on Earth and in the Ocean? Of those, only about 1.2 million have been formally described by scientists, meaning roughly 86 percent of land species and 91 percent of marine species have not yet been named or catalogued.
That gap matters for the extinction question in two ways. First, if the true number of living species is higher than we think, the extinction percentage goes up slightly because more species existing now still represents a tiny fraction of the cumulative total. Second, and more practically, species are going extinct today that science has never documented. We have no way to count them in the “extinct” column because we never knew they existed. This means both sides of the equation, the living and the dead, are underestimates. The ratio is probably roughly correct, but the absolute numbers behind it remain fuzzy.
Cryptic Species and Fossil Identification
Even the species we think we can count in the fossil record may be miscounted. Paleontologists identify fossil species primarily by the shape of their bones and shells, but modern genetics has revealed that many groups contain “cryptic species”: organisms that look nearly identical physically but are genetically distinct. Crocodylians, for example, have turned out to harbor multiple cryptic species within what were thought to be single widespread species, raising questions about how reliably we can identify distinct species in their fossil relatives.7Zoological Journal of the Linnean Society. Modern cryptic species and crocodylian diversity in the fossil record
Work on agamid lizards found that species-level diversity in the fossil record is likely underestimated, with around 30 to 40 percent of species pairs showing no significant shape differences when analyzed statistically. When researchers simulated the conditions of a fossil specimen, with incomplete bones and small sample sizes, classification accuracy dropped further.8PubMed. Are Modern Cryptic Species Detectable in the Fossil Record? A Case Study on Agamid Lizards The implication is that the fossil record likely lumps together species that were biologically distinct, which means the total number of extinct species is higher than fossil catalogs suggest. That pushes the overall extinction percentage upward, though by how much is anyone’s guess.
Are We in a Sixth Mass Extinction?
The debate over whether we are currently living through a sixth mass extinction event hinges on comparing modern extinction rates to the background rates established from the fossil record. One influential study estimated that the average rate of vertebrate species loss over the past century is up to 100 times higher than the background rate. Even using a conservatively high background estimate of 2 mammal extinctions per 10,000 species per 100 years, the number of vertebrate species that have disappeared in the last century would normally have taken between 800 and 10,000 years to vanish.9PubMed Central. Accelerated modern human-induced species losses: Entering the sixth mass extinction
Not everyone agrees with the mass-extinction framing. A recent analysis in Trends in Ecology and Evolution noted that while recent extinction rates do exceed fossil background rates, exceeding background rates is not by itself evidence of a mass extinction, because extinction rates have spiked above background at various points in Earth’s history outside the recognized Big Five events.10PubMed. Questioning the sixth mass extinction The distinction matters for how we frame the current crisis, but it does not change the practical reality: species are disappearing faster than they have in millions of years, and the losses are accelerating.
Whatever you call the current situation, it adds a thin but growing layer to the cumulative extinction percentage. The number of species lost to human activity is still small relative to the billions lost over geological time, so the headline figure stays at roughly 99 percent. But the rate at which that percentage is being padded has no precedent in recent geological history outside genuine mass extinction events.
Coextinction and the Hidden Losses
When people think about extinction, they tend to picture a charismatic animal like a dodo or a woolly mammoth. But the biggest category of extinction losses may be among organisms most people never think about: parasites, mutualists, and other species whose survival depends on a specific host or partner. When a host species goes extinct, its dependent parasites, commensals, and mutualistic partners face coextinction. Models suggest that coextinction may actually be the most common form of biodiversity loss.11PubMed Central. The sixth mass coextinction: are most endangered species parasites and mutualists?
The extinction of a single free-living species is rarely an isolated event. Dependent species face cascading risks as their hosts decline, and these cascades can ripple through ecological networks across trophic levels.12Annual Review of Ecology, Evolution, and Systematics. Coextinction and Persistence of Dependent Species in a Changing World Projections for parasitic worms under climate change, for instance, suggest that up to 30 percent of species may be committed to extinction from a combination of direct climate effects and the loss of their hosts.13PubMed Central. Parasite biodiversity faces extinction and redistribution in a changing climate
Coextinction is relevant to the overall percentage question because parasites and mutualists are enormously diverse. Every bird species, every mammal, every insect hosts its own community of species-specific parasites. If each vertebrate extinction drags several dependent species with it, the true toll of any extinction event, whether ancient or modern, is substantially higher than what the fossil record of free-living animals shows. This is yet another reason the 99 percent figure is almost certainly an undercount.
Extinction Debt and Species Already Committed to Disappear
One of the more unsettling concepts in conservation biology is extinction debt: the idea that habitat destruction or environmental degradation can set species on a path to extinction long before they actually vanish. A population can persist for years or decades in a degraded landscape, slowly declining until it drops below the threshold needed for long-term survival. Accumulating evidence suggests that these delayed extinctions are a significant and often unrecognized challenge across a wide range of organisms and ecosystems.14PubMed. Extinction debt: a challenge for biodiversity conservation
A study of 147 fragmented European grasslands found that the current diversity of long-lived plant specialists was better explained by historical landscape patterns than by the landscape as it exists today, meaning those plants were essentially living on borrowed time, their eventual disappearance already locked in by past habitat loss.15PubMed Central. Habitat fragmentation causes immediate and time-delayed biodiversity loss at different trophic levels The time lag between the trigger and the actual extinction can vary enormously, from as little as 5 years to as long as 570 years in observed cases, with projections suggesting some debts could take up to 1,000 years to fully settle.16Ecography. Understanding extinction debts: spatio–temporal scales, mechanisms and a roadmap for future research
Extinction debt means that the number of species currently alive overstates the number that will be alive in the near geological future, even if all destructive human activity stopped tomorrow. Some fraction of today’s living species are effectively already extinct in a functional sense; their populations have been reduced to the point where long-term viability is gone, and the final disappearance is a matter of when, not if. Species with long generation times and small, fragmented populations are the most likely to be carrying this kind of debt.
Why the Exact Number May Never Be Known
Every layer of the calculation pushes in the same direction: the true extinction percentage is almost certainly higher than the commonly quoted 99 percent, not lower. Soft-bodied organisms are undercounted in the fossil record. Cryptic species that look alike but are genetically distinct inflate the number of species that have existed without leaving distinguishable traces. Parasites and mutualists multiply the toll of every host extinction. Microbial diversity is so vast and so poorly understood that it could double or triple the total number of species that have ever lived without changing the number we can actually document.
At the same time, the living side of the fraction is also uncertain. With 86 percent of existing eukaryotic species still undescribed, the denominator is a moving target.1PLOS Biology. How Many Species Are There on Earth and in the Ocean? And extinction debt means some fraction of those undescribed species are already functionally doomed, just waiting out the lag. The whole calculation is built from estimates of estimates, each carrying its own wide confidence interval.
What scientists can say with confidence is that the living biota represents a thin surviving sliver of the total diversity that has existed on this planet. Whether that sliver is 1 percent, 0.1 percent, or 0.01 percent of the cumulative total depends on assumptions about microbial life and deep-time turnover that may never be resolvable with the data we have. The 99 percent figure serves as a useful shorthand for a real phenomenon: life on Earth is, and always has been, overwhelmingly a story of species that no longer exist.