What Nostradamus Actually Predicted — and What He Never Did
Verified biography, real quatrains, famous retrofits and outright fabrications — the complete, sourced record of what the physician from Provence actually wrote, and the one dated prediction that failed.
Nostradamus is not the father of prediction. He is the father of the unfalsifiable prediction — the kind that can never be scored, and therefore never wrong.
The man, minus the myth
The verifiable facts are these. Michel de Nostredame was born in December 1503 — the records permit either the 14th or the 21st — in Saint-Rémy-de-Provence, to a family that had converted from Judaism to Catholicism a generation earlier. He enrolled at the University of Avignon around 1519 but the university closed within about a year when plague swept the region. He then spent years traveling as an apothecary, and when he later sought a doctorate at Montpellier, the faculty expelled him almost immediately upon discovering that manual-trade past: apothecary work was forbidden to physicians by university statute, and the expulsion document survives in the university archives.
His plague credentials were real for their era. He worked outbreaks across southern France, promoting a rose-hip preparation and, by the standards of the day, unusually sensible hygiene. The era answered him with its own arithmetic: his first wife and their two children died in 1534, most likely of plague, while he was treating other people's families. He resettled in Salon-de-Provence in 1547, remarried, and raised six children there. From 1550 he published annual almanacs — astrological calendars thick with weather, harvest and political forecasts — and they, not the quatrains, made him famous and prosperous in his own lifetime.
The prophetic celebrity followed the almanacs. Catherine de' Medici, an avid consumer of the occult sciences, summoned him to court in 1555, and he drew up horoscopes for the royal children; in 1564 Charles IX named him a counselor and physician-in-ordinary. He died at Salon on the night of 1–2 July 1566, of complications his own writings suggest he saw coming — his secretary later reported he said he would not survive the night, an anecdote that is touching, unverifiable, and an early specimen of the genre that would engulf his memory: the story added after the fact.
Les Prophéties: what was actually published
The book that carries the legend appeared at Lyon in 1555, printed by Macé Bonhomme: a preface addressed to his son César, followed by 353 rhymed quatrains organized into Centuries — sets of one hundred verses. A 1557 edition extended the collection, and the canonical total of 942 quatrains across ten Centuries assembled itself only in omnibus editions printed after his death, in 1568 and later. Century VII contains just forty-two quatrains, and nobody has ever established why. There is no autograph manuscript; what we have are printed editions with variants, misprints and posthumous accretions — a textual situation closer to a rumor than a testament.
What do the verses say? Read without the highlight reel, they are a weather system of early-modern dread: floods, fires, plagues, famines, battles, betrayed cities, dying kings, monstrous births, comets. Nearly all are undated, unlocated or both. Critical scholarship — Pierre Brind'Amour's 1993 edition of the first Centuries, Peter Lemesurier's source studies — has shown that much of the material paraphrases earlier texts: the Mirabilis Liber anthology of prophecies, Roman historians, medieval chronicles. A significant fraction of the prophecies are, demonstrably, the past re-costumed as the future.
One more piece of the record deserves daylight, because his admirers never quote it: the almanacs. Unlike the quatrains, the almanac predictions were dated — tied to specific months of specific years — and therefore checkable. One modern count puts their total at 6,338 predictions. They performed the way sixteenth-century astrological weather and politics forecasting performed, which is why no one anthologizes them. The undated verse survived; the dated record did not. Keep that asymmetry in view for the rest of this article, because it is the entire mechanism.
The retrofitting machine: Hister, the young lion, and the art of aiming afterward
The most famous claimed hit is Hitler, and it dissolves on contact with a Latin dictionary. Three quatrains — II.24, IV.68 and V.29 — contain the word “Hister.” Hister is the classical Latin name of the lower Danube; Nostradamus, a man of his century's learned geography, used it as a river name, and quatrain II.24 speaks of beasts swimming across rivers and of the greater part of the battlefield being “against the Hister” — sensible for a river, strained for a man. The Hitler identification was popularized in the 1930s and 1940s, four centuries after publication, and required amending both spelling and grammar to work. Both wartime propaganda machines then ran with it: fabricated Nostradamus leaflets became a minor front of the war itself.
The oldest claimed hit — the one that allegedly launched the legend — is the death of Henri II. In July 1559 the king died of a lance splinter through the visor, received in a tournament joust. Quatrain I.35, printed in 1555, speaks of a young lion overcoming an old one on a field of combat in single battle, piercing his eyes in a golden cage. It reads like a bullseye — until you consult the timeline. As Lemesurier and other researchers established, no printed source connects that quatrain to the king's death until 1614, fifty-five years after the event. Neither contemporaries nor Nostradamus himself made the link; he favored a different verse, and even that retroactively. The most celebrated prophecy in the canon is a seventeenth-century discovery about a sixteenth-century accident.
The pattern generalizes. Napoleon and Hitler as the first two “Antichrists” of a supposed trilogy; the French Revolution; the atomic bomb; the moon landing; the death of Diana; every pandemic including the most recent one — in each case the procedure is identical. The event occurs. A reader with 942 vague verses and total interpretive license finds a match. The match is publicized as foresight. The direction of inference always runs from the present into the book, never from the book into the future. In forecasting terms this is postdiction, and its predictive value is exactly zero: no quatrain has ever been matched to a specific event before the event occurred.
The fabrication file: the 9/11 quatrain that Nostradamus never wrote
Within hours of the September 11 attacks, a quatrain flooded inboxes worldwide: two brothers torn apart by chaos, a great city falling, a third big war beginning. It was attributed to Nostradamus, sometimes with the date 1654. It appears in no edition of Les Prophéties, for a simple reason: it was written in 1997 by Neil Marshall, a student at Brock University in Canada, inside an essay arguing that Nostradamus's prophetic reputation is an artifact of vague language. Marshall composed the verse deliberately, as a demonstration that anyone can write lines abstract enough to seem prophetic after almost any catastrophe.
The internet then performed the second half of his demonstration for him, involuntarily and at planetary scale. Someone lifted the verse from the essay, someone else appended extra lines — the “two metal birds,” the “third big war” — and the composite circulated as a genuine sixteenth-century prophecy while the towers still burned. Fact-checkers, Snopes most prominently, traced the chain within days. The debunking is now old; the artifact remains the cleanest natural experiment ever run on prophecy culture: a verse everyone knew within weeks to be fake had already outperformed every genuine quatrain in apparent accuracy.
The same file contains gentler specimens. “Mabus,” from quatrain II.62 — Mabus then will soon die, and there will come a horrible undoing of people and animals — has been solemnly identified over the decades as Saddam Hussein, Osama bin Laden, and an assortment of politicians whose names can be misspelled backward with sufficient will. Each candidate dies or retires; the name migrates to the next. A prophecy whose subject can be reassigned indefinitely is not identifying anyone. It is a blank field with a deadline that never comes.
The one dated prediction that could fail — and did
Buried in the corpus is a genuine rarity: a quatrain with a date precise enough to be tested. X.72 declares that in the year 1999, seventh month, from the sky will come a great King of Terror, to bring back to life the great King of the Mongols. For decades this was the crown jewel of Nostradamus literature — the appointment with apocalypse, scheduled for July 1999. The month arrived. Nothing resembling the verse occurred. By the unforgiving standard the quatrain itself set, the flagship dated prophecy of the entire collection is a recorded miss.
The aftermath is more instructive than the miss. Rather than record a failure, the interpretive industry re-dated it (calendar drift), re-located it (a solar eclipse crossed Europe that August), re-translated it, or re-assigned it to September 2001. Scholars added a drier note: the earliest printings read “deffraieur” — a defrayer, one who pays costs — rather than “d'effrayeur,” of terror, meaning the terror itself may be a later transcription artifact. Either way the episode is complete: given one falsifiable prediction, the prophecy failed, and the interpretation apparatus absorbed the failure without cost.
That absorption is the tell. In a calibrated system, a miss is scored, published and paid for in reputation. In the Nostradamus system, a miss triggers reinterpretation, and the text's authority emerges intact — enhanced, even, by the fresh round of attention. A body of work that cannot lose authority through error is not making predictions in any meaningful sense. It is exercising a different technology altogether, which is the subject of the next section.
Why vague verse survives: unfalsifiability as a technology of authority
Consider what the quatrains' vagueness accomplishes, treated not as a flaw but as a design. A verse with no date can never expire. A verse with no location can be claimed by every country's catastrophe. A verse with symbolic actors — lions, eagles, cities of fire — can cast any era's protagonists. Each disaster that strikes anywhere, forever, is a fresh marketing event for the corpus, and no disaster that fails to strike can ever count against it. The expected reputational return of an unfalsifiable prophecy is strictly positive and compounds across centuries. Whether or not Nostradamus intended this — and his stated fear of persecution suggests deliberate clouding, whatever its motive — he built the optimal survival machine for a text.
The reader completes the circuit. Psychology has names for the components: the Barnum effect, by which people read specific personal meaning into statements engineered to fit anyone; confirmation bias, which files the apparent hits and discards the thousands of non-events; and hindsight bias, which makes every match feel inevitable once known. Sixteen generations of readers have supplied these services to the quatrains free of charge. The prophet's genius, if that is the word, was to write text that outsources all predictive labor to the future reader while retaining all credit.
This is why the honest answer to “how accurate is Nostradamus” is not a percentage. Accuracy is a property of scoreable statements, and the quatrains were constructed — or at least selected by survival — to be unscoreable. Against 942 verses, four and a half centuries of world events, and unlimited interpretive license, the number of retrofit matches is effectively guaranteed to be large, and it measures the ingenuity of interpreters, not the foresight of the author. His dated almanac predictions, the ones that could be graded, are precisely the part of his output that history graded and forgot.
The modern replacement: probability, falsifiability, public scoring
The scientific answer to the quatrain took four centuries to assemble, and its parts are unglamorous. Explicit probability, so that confidence is stated rather than implied. A defined window and an objective resolution criterion, so that the outcome is a fact rather than a debate. A falsifiability line, in the Popperian sense — a written description of the world in which the forecast is wrong. A sealed timestamp, so nothing can be edited after the fact. And public scoring, using the rule the meteorologist Glenn W. Brier published in 1950: the squared gap between stated probability and realized outcome, a penalty structure under which overconfidence is mathematically self-harming. Philip Tetlock's forecasting tournaments demonstrated the final piece — that under these constraints, forecasting skill is real, measurable and trainable.
This observatory exists to run that discipline in public, under the old prophet's name, as a standing correction of his method. THE CORE — the convergence engine behind the site — reads present conditions through six documented frameworks, from Kondratieff waves to the Dalio debt cycle, and outputs sealed probabilistic simulations: twenty-nine live projections, each with probability, uncertainty range, time window, trigger events, sources and a falsifiability condition. On 3 July 2026 a separate Calibration Ledger sealed ten predictions with hard resolve-by dates. All ten are PENDING. No track record is claimed, because none has been earned yet; the first public Brier scoring arrives in July 2027, and the misses will be published at the same size as the hits.
So, what did Nostradamus actually predict? In the checkable sense of the word: nothing that can be verified as foresight, one dated flagship that failed, thousands of dated almanac forecasts that history quietly graded down, and 942 undated verses that predict everything and therefore nothing. What he actually built was more interesting than prophecy — a mirror with a five-hundred-year warranty, in which every generation discovers its own catastrophes and calls the recognition foresight. The respectful response is not to keep polishing the mirror. It is to do what he never did: write the number down, name the date, sign the failure condition, and stand there while history marks the work.
Frequently asked
Did Nostradamus predict 9/11?
No. The “two brothers torn apart by chaos” quatrain circulated after the attacks appears in no edition of Les Prophéties. It was written in 1997 by Neil Marshall, a Brock University student, as a deliberate demonstration of how vague verse can be made to look prophetic; extra lines were added as it spread. Fact-checkers traced the fabrication within days.
Did Nostradamus predict Hitler?
No. The word in quatrains II.24, IV.68 and V.29 is “Hister” — the classical Latin name of the lower Danube river, standard learned geography in the sixteenth century. The identification with Hitler was invented in the 1930s and 1940s, four centuries after publication, and requires altering the word itself.
Did Nostradamus predict the end of the world?
No. He gave no doomsday date for our era; the preface to Les Prophéties states his prophecies extend to the year 3797, a horizon, not an apocalypse appointment. His one famous dated verse — the seventh month of 1999 — passed without event, and later end-dates attributed to him are inventions of interpreters.
Are there real Nostradamus predictions for the future?
Not in any testable sense. The quatrains are undated and unfalsifiable, so none can be assigned to a specific future year except by a modern interpreter's choice. Genuinely forward-looking statements require explicit probabilities, deadlines and failure conditions — the calibrated method this observatory practices, with sealed projections scored publicly from July 2027.
Sources
- Nostradamus, Les Prophéties, Lyon: Macé Bonhomme, 1555; expanded 1557; posthumous omnibus editions (from 1568) totaling 942 quatrains
- Pierre Brind'Amour, Nostradamus astrophile: les astres et l'astrologie dans la vie et l'oeuvre de Nostradamus, University of Ottawa Press, 1993
- Peter Lemesurier, The Unknown Nostradamus, O Books, 2003 — including the finding that quatrain I.35 was first linked in print to the death of Henri II in 1614
- Edgar Leoni, Nostradamus and His Prophecies, 1961 (standard English reference edition of the full text)
- Snopes fact check, 'Did Nostradamus Predict the 9/11 Attacks?' — tracing the fabricated quatrain to Neil Marshall's 1997 Brock University essay
- Museum of Hoaxes, 'Nostradamus Predicted 9/11' (September 2001)
- Britannica, 'Nostradamus' — biographical entry (birth December 1503, Saint-Rémy-de-Provence; death July 1566, Salon)
- Glenn W. Brier, "Verification of Forecasts Expressed in Terms of Probability", Monthly Weather Review, 1950
- Karl Popper, Conjectures and Refutations, 1963 — falsifiability as the demarcation of testable claims
- Philip E. Tetlock & Dan Gardner, Superforecasting, Crown, 2015