In 1590, the governor of England's first American colony returned to Roanoke Island after three years away to find every one of his 115 colonists gone. No bodies. No signs of violence. No burned buildings. The only clue was a single word carved into a wooden post: CROATOAN. John White had left behind his own daughter, his son-in-law, and his infant granddaughter — the first English child born in the Americas. He never found any of them. More than four centuries later, neither has anyone else.
The Lost Colony of Roanoke is the oldest unsolved mystery in American history. It predates the United States by nearly two centuries, predates any stable English presence in North America, and has resisted every attempt at resolution despite being one of the most investigated disappearances in the archaeological record. It is a mystery that has generated more theories than evidence, more certainty than proof, and more cultural weight than almost any other colonial-era event — in part because it happened at the very beginning, when everything about the English presence in the New World was still uncertain, and the disappearance of an entire community was not yet impossible to imagine.
What happened at Roanoke
The Roanoke Colony was England's second attempt to establish a permanent settlement on the coast of present-day North Carolina. The first attempt, in 1585, had failed when the settlers — soldiers and gentlemen rather than farmers and tradespeople — ran out of food and were evacuated by Sir Francis Drake after a difficult winter. The second attempt, in 1587, was different in design: 115 men, women, and children, intended as genuine settlers rather than explorers, under the governorship of John White, an artist and cartographer who had been on the first expedition.
Almost immediately, the colony was in trouble. Relations with local Indigenous peoples were hostile — a legacy of violence from the first expedition that the new settlers inherited without having caused it. Supplies were inadequate. White was persuaded by his colonists to return to England to request resupply, leaving in late 1587 with the intention of returning within months. His granddaughter Virginia Dare, the first English child born in America, was nine days old when he left.
He did not return for three years. The Anglo-Spanish War of 1588 — the year of the Spanish Armada — meant that every available English ship was required for the war effort. When White finally reached Roanoke in August 1590, the settlement was abandoned. The buildings had been dismantled — not destroyed, but carefully taken apart, suggesting planned departure rather than emergency flight. The word CROATOAN was carved on a post. The letters CRO were carved on a nearby tree. There was no cross carved above the word — a distress signal White had arranged with the colonists before leaving. Whatever had happened, they had not considered it an emergency requiring rescue.
| Clue | What it suggests | What it doesn't explain |
|---|---|---|
| CROATOAN carved on post | Colonists moved to Croatoan Island (present-day Hatteras Island) or joined the Croatoan people | Why no colonists were ever found there; why contact was never re-established |
| CRO carved on tree | Confirms the Croatoan destination; carved in haste or as a waypoint marker | Why the full word wasn't completed |
| No distress cross carved | Departure was planned and voluntary, not an emergency | What the plan was and who organized it |
| Buildings dismantled, not burned | Orderly departure with time to prepare; materials taken or left for later use | Where the materials went; how 115 people moved without leaving a trail |
| No bodies found | No massacre; no epidemic that killed them in place | Where they went after leaving; why none were ever recovered |
| Chests and belongings left behind | Departure was faster than fully planned, or some goods were deliberately abandoned | Whether departure was entirely voluntary or partly forced |
The theories, ranked by evidence
Integration with the Croatoan people
The most evidence-supported theory holds that the colonists moved to Croatoan Island — present-day Hatteras Island — and were absorbed into the Croatoan Indigenous community. The Croatoan were the one local group that had maintained friendly relations with the English. The carved message was the agreed signal for relocation. Subsequent English expeditions, which were supposed to search Croatoan Island, were prevented from doing so by weather and by competing priorities. The colonists were simply never looked for properly.
Supporting this theory: accounts from the early seventeenth century describe Indigenous people in the region with gray eyes and European features. The First Colony Foundation, which has conducted the most systematic archaeological work on the Roanoke question, has found artefacts consistent with English origin at sites associated with the Croatoan people on Hatteras Island. DNA studies of families in the region with claimed Croatoan ancestry have produced suggestive but not conclusive results.
Dispersal into multiple Indigenous communities
A variant theory holds that the colonists did not stay together but dispersed into several different Indigenous communities across the region. Early seventeenth-century accounts from the Jamestown colonists — established in 1607, just 100 miles from Roanoke — describe reports of people with European features living among various Indigenous groups. John Smith recorded accounts of a community to the south where people lived in two-story stone houses — a description that fits neither Indigenous nor English building traditions of the period but might fit a mixed community. Chief Powhatan allegedly told the Jamestown colonists that he had destroyed a group of people who had been living with a neighboring tribe — people who, some historians argue, may have been the surviving Roanoke colonists or their descendants.
Spanish attack
Spain considered the entire North American coast to be Spanish territory and had intelligence about the English settlement at Roanoke. Spanish naval records document expeditions specifically intended to locate and destroy the English colony. Some historians argue that a Spanish attack could explain the disappearance — though the absence of bodies and the orderly dismantling of buildings argue against a violent end. The US National Archives holds Spanish colonial records that have been partially analyzed for references to Roanoke, so far without definitive results.
Starvation and dispersal
Tree ring data from the region, analyzed by climatologists studying the period, shows that 1587 to 1589 was one of the most severe droughts in the area for 800 years. The colony arrived with inadequate supplies and into a landscape that could not support them. In this theory, starvation drove the colonists to seek food with neighboring Indigenous peoples, and dispersal followed naturally from survival need rather than from any single decision or event.
What archaeology has and hasn't found
Archaeological investigation of the Roanoke question has been limited by a fundamental problem: no one is certain exactly where the 1587 settlement was located. The island has changed shape significantly over four centuries of erosion, and the most likely location of the original settlement may now be underwater.
A 2012 analysis of a John White map held by the British Museum revealed a previously hidden symbol beneath a patch — a four-pointed star — at a location about 50 miles inland from the coast. The British Museum and the First Colony Foundation interpreted this as a possible marker for a secondary site, perhaps a planned inland retreat. Subsequent excavation at the marked location — a site called Site X near Merry Hill, North Carolina — has produced English-period artefacts including pottery fragments consistent with early colonial manufacture. The excavation is ongoing.
The curious connection
The Lost Colony of Roanoke has been generating theories for four centuries, and the theories have always reflected the concerns of the era producing them. In the nineteenth century, when American national identity was being constructed around narratives of frontier conquest, the dominant theory was Native American massacre — the colonists as innocent victims of savage violence. In the twentieth century, as American attitudes toward Indigenous peoples shifted, the integration theory became dominant — the colonists as successful participants in cross-cultural survival. In the twenty-first century, climate data has made the drought theory newly compelling in an era acutely conscious of environmental collapse.
The mystery has not changed. The theories have changed because the people producing them have changed. What we find in the silence of Roanoke is not evidence of what happened — it is a screen onto which each generation projects its own understanding of how civilizations end.
This is not unique to Roanoke. Every unsolved historical mystery functions as a projective surface — a gap in the record that invites interpretation, and whose interpretations reveal more about the interpreter than about the mystery itself. The Voynich Manuscript tells us about our desire for hidden knowledge. The Nazca Lines tell us about our difficulty accepting that ancient people had purposes we cannot reconstruct. Roanoke tells us about our anxiety regarding cultural contact, survival, and the fragility of the social structures we build in hostile environments.
The colonists of Roanoke did not vanish. They went somewhere, did something, and became something. The record of what that was has been lost — partly through neglect, partly through the destruction of Indigenous knowledge systems that might have preserved it, and partly through four centuries of looking for the wrong things in the wrong places. The mystery is not that they disappeared. The mystery is that we have been unable, or unwilling, to find where they went.
Virginia Dare, nine days old when her grandfather sailed away, would have grown up somewhere. The question of where she grew up is a question about what America was before it decided what it was going to be.
FAQ
What happened to the Lost Colony of Roanoke?
The 115 colonists of the 1587 Roanoke Colony disappeared between Governor John White's departure in 1587 and his return in 1590. The only clue left was the word CROATOAN carved on a post and the letters CRO on a tree. The most evidence-supported theory holds that the colonists relocated to Croatoan Island and were absorbed into the Croatoan Indigenous community, but no definitive evidence has been found and the question remains officially unresolved.
What does CROATOAN mean?
Croatoan was the name of both an island (present-day Hatteras Island) and the Indigenous people who lived there — the only local group that had maintained consistently friendly relations with the English colonists. Governor White had arranged before his departure that if the colonists relocated, they would carve their destination on a post, and carve a distress cross above it if they were in danger. The absence of a cross indicates the departure was not an emergency.
Was Virginia Dare ever found?
Virginia Dare — born August 18, 1587, the first English child born in the Americas — was never found. She disappeared with the rest of the colony. She has become a significant figure in American cultural mythology, appearing in novels, poems, and legends across four centuries, sometimes as a symbol of the lost promise of English America and sometimes as a figure who survived and integrated into Indigenous life. No historical or archaeological evidence of her fate has been confirmed.
Why didn't John White search Croatoan Island when he returned in 1590?
White's return voyage was plagued by bad weather and the competing priorities of the ship's captain, who had his own commercial objectives. When White arrived at Roanoke in August 1590 and found the CROATOAN carving, he intended to sail immediately to Croatoan Island to search for the colonists. A severe storm prevented the search, damaged one of the ships, and the captain refused to continue. White never returned to America. The colonists were never formally searched for at Croatoan Island.
What has recent archaeology found about Roanoke?
A 2012 analysis of a John White map at the British Museum revealed a hidden symbol suggesting a secondary site about 50 miles inland. Subsequent excavation at this location — Site X near Merry Hill, North Carolina — has produced English-period pottery fragments consistent with early colonial manufacture. The First Colony Foundation continues excavation at multiple sites. No definitive proof of the colonists' fate has been found, but the inland site theory has gained significant archaeological support.