Digital Map Uncovers 300,000 km of Roman Roads, Shedding Light on Millennia of Archaeological Ambiguity

Across Europe, North Africa and the Near East, Roman roads structured mobility, the economy and administration for centuries. However, despite their fundamental role in imperial history, their exact route remained largely unknown. A consortium of researchers from the University of Aarhus (Denmark), the ICAC of Tarragona and the CNRS, published a new digital cartography in the journal Scientific Data: Itiner-e.

This project, based on analysis of historical sources, topographical maps, archaeological data and satellite imagery, reveals that Rome's road network spanned nearly 300,000 kilometers — almost double previous estimates. But above all it exposes a disturbing reality: less than 3% of these routes are known with precision. An essential map, therefore, but also an observation of the scale of the areas still to be documented.

A digital map to rebuild knowledge of Roman roads

Until recently, knowledge of the roads of the Roman Empire was based on limited resources. And most often fragmentary and unevenly distributed. The main references, such asAtlas Barrington or the Digital Atlas of Roman and Medieval Civilizations (DARMC), offered global mapping. But it remained of low resolution, poorly adapted to local geographical and archaeological realities. With the project
Roamthis limit has now been largely exceeded.

Launched between 2020 and 2024, Roam is the result of a collaboration between more than twenty researchers from institutions such as the University of Aarhus (Denmark), the ICAC (Tarragona, Spain), the CNRS (France) and the University of Leiden (Netherlands). The objective: to constitute an open, precise and georeferenced database of the Roman road network as it existed around 150 AD, at the territorial apogee of the Empire.

© Itiner-e

Newly established map of Roman roads.

This work mobilized thousands of crossed sources: ancient texts like theAntonin's Itinerary or the Puisinger tableregional archaeological publications, epigraphic databases (such as READ for mile markers), satellite images (Google, ESRI), historical aerial photos, 19th century topographical maps and data from the Corona mission (military satellite photography from the Cold War). The researchers digitized each section of road into a GIS (geographic information system). They thus took into account the real topography — for example, by following passes rather than theoretical straight lines.

Result: a digital atlas of 299,171 km of roads, divided into 14,769 segments. A spectacular jump compared to the 188,555 km mapped in previous datasets. This cartographic overhaul places the Roman road at the heart of imperial dynamics.

A structured network, but unevenly documented

One of the major contributions of the project Roam lies in the categorization of the roads and the spatial precision of the segments. Researchers distinguish two types of axes: main roads and secondary roads, according to archaeological and textual criteria. The first, documented by mile markers or ancient sources, represent 103,478 km (or 34.58% of the total). These roads served major urban and military centers, reflecting the Empire's desire for administrative and logistical unification.

© Itiner-e

Image of a recently discovered milestone in Cilicia, Turkey.

Conversely, secondary roads, often local or regional, cover 195,693 km, or 65.42% of the network. Their documentation remains more incomplete, because they rarely left visible traces or epigraphic mentions. However, their importance is not negligible. They reflect the logic of fine connectivity between villages, farms, mines, quarries, inland ports, etc. Their inclusion in the map underlines the complexity of the Roman territorial network, far from the image of a purely hierarchical network.

But this coverage remains uneven depending on the geographic areas. Certain regions, such as the Iberian Peninsula, Greece or Tunisia, benefit from abundant documentation resulting from decades of research. Others, such as central Anatolia, the interior Balkans or Dacia, remain more hypothetical, due to lack of excavations or precise sources. This heterogeneity is reflected in the “trust” map (confidence map) developed by the authors. It combines road density and source reliability.

Finally, cutting into segments allows for detailed analysis. Each section is accompanied by precise metadata — geodesic length, average slope, certainty of location, bibliography, URI — facilitating its integration into multidisciplinary research (history, archaeology, geography, modeling).

Spatial uncertainty widely assumed

One of the most striking findings from the project
Roam is the degree of uncertainty surrounding the exact location of roads. Of the 299,171 km identified, only 2.737% are classified as “certain” according to spatialization criteria. This means that their route is documented with an accuracy of less than 50 meters in mountainous terrain and 200 meters in the plains. These segments generally come from archaeological excavations, field observations or very detailed photographs.

Archaeologists therefore consider the majority of the network (i.e. 89.818%) as “conjectured”. In these cases, the routes were identified based on coherent clusters of clues: alignment between known ancient sites, tracking of ancient paths visible on old maps or satellite imagery, logical topographical continuities. These segments are certainly plausible, but without direct proof. Finally, 7.445% of the network is qualified as “hypothetical”, designating the supposed routes, but for which the material evidence still remains insufficient.

This gradation constitutes one of the great methodological strengths of the project. It allows a critical reading of the map and encourages caution in analyses. Additionally, it specifically identifies areas that merit future research. As Tom Brughmans explains to LiveScience“ This project is also a call to the scientific community to fill these gaps “.

Another important consequence: this uncertainty is historically instructive. It reflects the evolving nature of the road network. Many Roman roads followed previous routes, sometimes even prehistoric, modified over the centuries. Some have been abandoned, others integrated into modern roads. This documentary blur reminds us that the road is not like a fixed object. It is an infrastructure in perpetual reinvention.

An open tool for multidisciplinary research

Beyond cartography, Roam is intended to be a scalable work platform. Accessible online on itiner-e.org, it allows you to consult, download and analyze all data under open licenses. Each section has a unique identifier (URI), and is linked to bases such as Pleiades (antique gazetteer) or
Vici.org (archaeological atlas). This interoperability makes it a powerful base for digital studies.

There are many potential uses. In economic history, roads make it possible to model the flow of goods, to estimate transport times or to reconstruct commercial circuits. In archaeology, the map facilitates the identification of passage zones, strategic intersections or regions neglected by excavations. In historical epidemiology, Roam can help study the spread of pandemics such as the Antonine plague (2nd century), which is said to have killed up to a quarter of the Roman population.

The authors also emphasize the educational interest of the project. By visualizing the topographical reality of journeys, we better understand why roads avoid certain areas or follow specific valleys. We also see how the relief, borders or natural resources have structured the connections. The myth of the Roman straight line is thus shattered.

Finally, Roam is thought of as a constantly updating project. Researchers can suggest additions or corrections. This collaborative logic, coupled with complete methodological transparency, makes it a sustainable tool. Above all, this map becomes a compass for future research.

Source: de Soto, P., Pažout, A., Brughmans, T. et al.“Itiner-e: A high-resolution dataset of roads of the Roman Empire”. Sci Data 12, 1731 (2025).