Manhattan's 138 lost cemeteries: a construction risk hidden under New York's streets

*Figure 1. A version of the Castello Plan of New Amsterdam (1660) as published by Joan Blaeu, one of the historic maps layered into the NYU team's GIS. Source: Laefer et al. (2026), TC301 Athens.*

An NYU team has mapped burial grounds that have disappeared from Manhattan's maps. They argue that a shared, GIS-based record of forgotten graves should become a standard pre-construction tool, so that projects are not derailed by unexpected human remains.

Key facts

158 cemeteries identified, the earliest from the mid-17th century; only 20 survive with visible grave markers and 138 are lost
The African Burial Ground, found in 1991 during preparation for two federal buildings, held more than 15,000 intact skeletons 25 to 30 feet below ground on a 6 acre site
That discovery caused a multi-year delay, the cancellation of one of the two planned buildings and the creation of a national monument, and scuttled a $54 million federal investment
10 seventeenth-century burial sites were found in lower Manhattan
Manhattan burial bans: south of Canal Street in 1832, extended to 14th Street in 1839 and to 86th Street in 1852

For owners and contractors, few site discoveries are as disruptive as human remains. Such a find usually delays a project by years and brings cost overruns nobody planned for. A paper presented at the 4th International Symposium on Geotechnical Engineering for the Preservation of Monuments and Historic Sites (ISSMGE TC301) in Athens argues that in older American cities this is a risk that can be foreseen and managed. Debra F. Laefer, Professor at New York University, and colleagues at NYU (J. Almary, A. Rozhkov, M. Yeoh, F. Tyner, S. Tyazhelkov, E. O'Sullivan and D. Doherty) have been reconstructing Manhattan's forgotten burial grounds. They count more than 138 lost burial locations in that borough alone, and they propose a shared digital resource to stop them being dug up by accident.

How cemeteries get lost

The authors explain that repeated colonization, industrialization and large waves of immigration have kept reshaping New World cities: their communities, their institutions and their subsurface. Immigrant and rural communities built churches, cemeteries and paupers' graveyards. Later generations moved to wealthier neighborhoods and often left their ancestors behind. Over time, many burial grounds vanished from local maps, frequently with the bodies still in the ground.

Standard site investigation has trouble detecting them. Geophysical methods are limited in the depth they can reach, asphalt and concrete paving get in the way, and centuries of changes to street layouts and ground levels make it hard to predict where anything is buried. The authors point to nearly 200,000 reported utility strikes each year as evidence of how incomplete subsurface records are, and note that finds of human remains are not uncommon.

Regulation added to the problem. Growing populations and repeated epidemics led New York to restrict burials in stages: south of Canal Street in 1832, up to 14th Street in 1839 and up to 86th Street in 1852. Burial in Manhattan was eventually prohibited altogether. The Rural Cemetery Act of 1847 allowed private companies to open large cemeteries far from populated areas, mostly on what is now the Brooklyn and Queens border. According to the paper, these policies helped erase collective memory of the older burials.

Figure 2. Manhattan's 19th-century burial restriction zones (south of Canal Street 1832, 14th Street 1839 and 86th Street 1852) shown against the 17th-century Dutch colony and the Land of the Blacks, which stretched about 3 km north. Source: Laefer et al. (2026), TC301 Athens.

The cost of not knowing

The project identified 158 cemeteries, the oldest dating from the middle of the 17th century. Only 20 still exist as sites with visible grave markers. The other 138 are lost, and very few have been rediscovered. The best known rediscovery is the African Burial Ground, a 6 acre site found in 1991 while two new federal buildings were being prepared. More than 15,000 intact skeletons were found 25 to 30 feet (roughly 8 to 9 m) below ground. The discovery caused a multi-year delay and the cancellation of one of the two buildings, and the site became a national monument.

The authors stress that the 1991 discovery did not happen for lack of information: facts and traces of the burial ground appeared in various reports and maps. The find ultimately scuttled a $54 million federal investment which, they argue, could have been completely avoided. Other cases show the same pattern. The Old Graveyard of New Amsterdam, regarded as the colony's earliest non-native burial ground and already described as wholly in ruins in 1656, was divided into lots and sold on British orders in 1676. No reinterment is recorded, and workmen found human remains there while digging cellars in the mid-19th century. The French Protestant cemetery, founded in 1688, had its graves removed after the church was sold in 1831, and a skyscraper now stands on the site.

Figure 3. Lost and existing cemeteries in Manhattan by religious affiliation, coloured by date. The inset shows lower Manhattan, with the Old Graveyard of New Amsterdam, Trinity Church Cemetery, First Shearith Israel Cemetery and Trinity Lutheran Cemetery highlighted. Source: Laefer et al. (2026), TC301 Athens.

Rebuilding the map of New Amsterdam

The team focused on lower Manhattan between 1633 and 1766, from State Street in the south to the line of today's Vesey, Ann, Beekman and Pearl Streets. The aim was to understand in depth how cemeteries are lost and how they can be found again. The project brings together sources that were available but had never been combined. These include the Castello Plan of 1660, the Montresor Plan of 1766 and the Ratzer Map of 1776, Dutch documents held by The Holland Society of New York, local burial legislation, and published compilations of burial sites.

The team's method uses the link between where people lived, where they worshipped and where they were buried. Congregants were identified from birth, marriage and death records and then matched to property deeds in the six-volume Iconography of Manhattan Island. This allowed a religious affiliation to be assigned to the owners of houses and businesses. The data were built into a GIS, with each layer covering one theme. For 1633-1660, the team identified 98 named residents with a known religious affiliation and 95 without one. Congregants tended to live in rough rings around their house of worship. The team used the member living furthest from each church to define its sphere of influence, a method that can help narrow down where an associated burial ground is likely to be.

Figure 4. GIS layer over the Castello Plan, geo-registered to the current New York City base map. Shaded areas show the spatial extent of each congregation around its house of worship. Source: Laefer et al. (2026), TC301 Athens.

The work also recovers communities that are often missing from the official record. Black residents were barred from burial in the formal graveyards. North of the Dutch wall, in what is now Greenwich Village, lay the Land of the Blacks: around 30 to 40 farms, most of them owned by half-free Black people between 1643 and the early 1700s. Using individual land deeds, the team identified four Black landowners who do not commonly appear in other sources. In total, the team found 10 seventeenth-century burial sites in lower Manhattan.

Lessons for practice

Treat historic burial grounds as a recognised subsurface hazard in older cities, alongside buried utilities, on private land and in the public right of way.
Documentary and demographic evidence, such as congregations, land ownership and burial laws, can narrow down likely burial locations where geophysics struggles beneath paving.
Much of the evidence already exists. The problem is that it is scattered, and bringing it together in a GIS is the key step.
The authors see combined large language and vision models (such as LLaVA) as a way to reduce the labour of finding and processing historic sources, making the approach scalable and machine-readable.

The authors' longer-term goal is a shared, city-wide, GIS-based pre-construction resource that is automatable, AI-ready and transferable to other North American cities. For infrastructure owners, it could turn an unpredictable, project-stopping find into a risk that is identified at the desk study stage.