The Lost Colony of Roanoke: What the Evidence Actually Says

Engraving after John White showing English colonists arriving in Virginia in the 1580s, from the era of the Roanoke voyages

The Englishmen's arrival in Virginia — engraving by Theodor de Bry after John White, 1590. Public domain, via Wikimedia Commons.

On 18 August 1590, Governor John White finally stepped back onto Roanoke Island, off the coast of present-day North Carolina. It was his granddaughter's third birthday. Three years earlier he had sailed for England leaving behind more than a hundred English colonists — men, women and children, including his own daughter Eleanor Dare, her husband Ananias, and their baby, Virginia Dare, the first child born to English parents in the Americas. He expected a struggling but living settlement. He found the houses taken down, the site enclosed by a new palisade, and nobody there at all.

What he did find has fuelled America's oldest colonial mystery for more than four centuries: the letters "CRO" carved into a tree, and one word carved in capitals on a post of the palisade — CROATOAN. No bodies. No battle signs. No cross — and that matters, as we will see. Popular retellings stop there and invite ghosts. This article follows the evidence — White's account, tree rings, maps and modern archaeology — and separates what is established, what is probable, and what is not known.

England's first attempt: 1585–1586

Roanoke was Sir Walter Raleigh's project. Backed by a royal patent from Queen Elizabeth I, Raleigh sponsored England's first attempts to plant a permanent base in North America — a foothold against Spain and a hoped-for source of wealth. The first expedition, commanded at sea by Sir Richard Grenville, established a colony on Roanoke Island in July 1585 under Governor Ralph Lane. John White sailed with it as artist and mapmaker, producing the watercolours of Algonquian people, villages, plants and wildlife that remain our most informative visual record of the Native society of that coast.

Lane's colony was a military outpost of about a hundred men, dependent on food traded — and eventually fought over — with the local Secotan and neighbouring peoples. Relations collapsed into violence, supplies ran out, and in June 1586 the colonists abandoned the island with Sir Francis Drake's passing fleet. Grenville, arriving later with supplies, left a holding party of fifteen men. England would try again — this time with families.

1587: the colony of families, and a departure that became three years

On 7 January 1587, Raleigh chartered a new venture — the "Cittie of Raleigh" — with John White as governor. In May 1587 White sailed with more than a hundred colonists: around 115 to 117 people in most reconstructions (published totals range from about 112 to 121, which is why sources differ), including seventeen or so women and, for the first time, children. This was meant to be a self-sustaining farming settlement, and its intended destination was actually the Chesapeake Bay, not Roanoke. The expedition's pilot insisted on stopping at Roanoke to check on the fifteen men left behind — they found the fort razed and evidence that the holding party was gone — and then refused to carry the colonists further. The families settled on Roanoke by default, in a place already known to be short of food and on bad terms with some of its neighbours.

There were bright moments. Manteo, a Croatoan leader who had travelled to England and become the colonists' essential ally and interpreter, was baptised and named Lord of Roanoke. On 18 August 1587, Eleanor Dare gave birth to Virginia Dare. Nine days later, the colonists persuaded White to return to England for supplies — meant to be a quick trip. It became neither: England was at war with Spain, the Armada came in 1588, shipping was requisitioned, and White's attempts were blocked or turned back. He finally reached Roanoke on 18 August 1590.

John White's watercolour of the Algonquian village of Secotan, showing its houses, fields and temple

The village of Secotan, watercolour by John White, c. 1590 — public domain via Wikimedia Commons.

What White found — and what he did not find

White's own narrative of the 1590 voyage is the primary source for everything that follows, and its details are more orderly than the legend suggests. The settlement had not been burned or sacked. The houses had been dismantled — "taken downe" — and the site had been newly fortified with a high palisade, suggesting planned work over time, not a midnight flight. Heavy items too bulky to carry casually had been buried or left in an organised way. There were no graves of a hundred dead, no bones, no sign of a massacre at the site.

Above all, there were the carvings, and the agreement behind them. Before leaving in 1587, White and the colonists had arranged a signal: if they moved away, they would carve the name of their destination on a tree or post; and if they had been forced to move — by attack or duress — they would carve a Maltese cross above the name. "CRO" was cut into a tree near the shore; "CROATOAN" was cut into a palisade post, in fair capital letters. No cross was found with either carving. By the colonists' own agreed code, they had moved voluntarily, and they had said where: Croatoan — the island and people of Manteo's community, roughly modern Hatteras Island, some fifty miles south.

White believed them. He intended to sail to Croatoan at once. Weather destroyed the plan: a storm drove his ships, an anchor was lost, crews refused further risk in worsening conditions, and the expedition turned for England. White never returned to America. He wrote his account of the final voyage in February 1593 and died around that year. No English ship is known to have checked Croatoan in time to matter. That failure of follow-up — not a supernatural vanishing — is the true hinge of the mystery. The colonists left a forwarding address, and nobody went there.

The evidence since: drought, maps and spades

Four lines of modern evidence have reshaped the story, and none of them is a ghost story.

  1. The worst drought in 800 years. In 1998, researchers led by David Stahle of the University of Arkansas published a tree-ring study in the journal Science, reconstructing climate from ancient bald cypress trees in the region. Their finding: 1587–1589, the exact years the Lost Colony was on its own, was the most extreme three-year drought in about 800 years of reconstructed regional climate. For a small settlement dependent on its own crops and on trading for Native-grown corn — corn the Secotan and Croatoan also desperately needed in a drought — this is a powerful pressure toward dispersal, dependence on Native hosts, or starvation. It does not prove any single fate, but no serious account of Roanoke can now ignore it.
  2. The map with the hidden fort. In 2012, researchers examining White's map La Virginea Pars at the British Museum found a small patch pasted over part of the map, concealing a fort-like symbol about fifty miles inland, near the Chowan River in present-day Bertie County — upriver, fertile, and away from exposed Roanoke. A map patch is not proof that colonists went there, but it is a contemporary clue, not a modern invention.
  3. Archaeology inland — "Site X." First Colony Foundation excavations in Bertie County, at the area dubbed Site X, have recovered late-sixteenth-century English material, including Surrey-Hampshire Border ware pottery of the right period, in a Native settlement context. The quantities are small — consistent with a small group of English people living within or beside a Native community, not a transplanted town. This suggests some colonists may have lived there; it does not identify them by name.
  4. Archaeology on Hatteras. On Hatteras Island — Croatoan itself — excavations have found English objects of the right era, including a rapier hilt and gun hardware, mixed into Native contexts. Objects travel by trade, so care is needed — but English goods of the 1580s pattern, in the very place the carving named, are consistent with colonists reaching Croatoan and being absorbed into that community over years.

Weighing the theories honestly

So what happened? Ranked by evidence, not by drama:

  • Dispersal and assimilation (best supported). The colonists, hit by drought and short of food, split up — exactly what later Jamestown-era evidence and Algonquian geography would predict. Some went to Croatoan/Hatteras; another party may have moved inland toward the Chowan. Over years, survivors integrated into Native communities. This fits the carvings, the missing distress cross, the dismantled houses, and both archaeology sites. It remains unproven in its details — no grave or artefact has yet been tied definitively to a named colonist.
  • Starvation and disease (plausible, partial). The drought makes hunger highly likely, and some colonists surely died in those three years. But starvation alone does not explain the orderly dismantling or the carving.
  • Attack or massacre (weakly supported). Spanish attack was feared — Spain did reconnoitre the coast — but no Spanish record claims a destruction of Roanoke, and White found no evidence of violence at the site. Violence in later dispersal cannot be ruled out; a single annihilating attack is not what the site showed.
  • The Dare Stones and other dramatic "solutions" (not accepted). In the 1930s–40s, a series of carved stones purporting to tell Eleanor Dare's story surfaced in the Southeast. They are regarded by historians as hoaxes, with the possible exception of the first stone, whose authenticity is still disputed but unproven. They are evidence about the twentieth century's appetite for the mystery, not about 1590.

The National Park Service, which cares for Fort Raleigh National Historic Site, still states the position plainly: the fate of the Lost Colony is unresolved. That is history working correctly. "Unresolved" no longer means "vanished without a trace" — the traces point south to Croatoan and west up the rivers, and the spades are slowly following them.

Why Roanoke still matters

Roanoke failed as a colony and succeeded as a beginning. Its maps and White's watercolours shaped Europe's picture of America, and its families were the template Jamestown would follow twenty years later. Its mystery, properly told, is not about people evaporating. It is about 117 people who, struck by the worst drought in eight centuries, made a rational decision, carved their destination for the man who had promised to return — and waited for a rescue that weather, war and one lost anchor turned aside at the final mile.

References

Fact notes: Colonist totals are a range (c. 112–121; NPS standard 117) because reconstructions differ. Bertie County/Hatteras finds are "consistent with," not proof of, colonist presence. Dare Stones: regarded as hoaxes.

Comments