Roman roads still shape how we move through the landscape. What if some of the routes shaping modern travel were chosen by Roman surveyors nearly two thousand years ago? A new study has built the most detailed empire-wide digital model yet of the Roman road system, then tested assumptions that have survived in textbooks for generations.
The result is more interesting than the familiar image of roads running perfectly straight from Rome. Roman builders did favour direct routes when the landscape allowed it, but hills, valleys and existing settlements forced them to adapt. Their network was practical, uneven and deeply tied to the terrain.
A map of 299,171 kilometres
The research team assembled a spatially detailed dataset covering 299,171 kilometres of roads across the Roman Empire. At its greatest extent, the empire connected much of Europe, North Africa and the Middle East. Putting those routes into one model allowed the researchers to compare roads across provinces instead of treating each surviving stretch as an isolated monument.
Keep exploring
Continue with Curiosmos
That matters because the surviving evidence is uneven. Some roads are preserved as visible paving, bridges or milestones; others survive only as mapped traces. The study is a quantitative exploration of the network, not a claim that every ancient kilometre has been recovered. Its value lies in making the limits and patterns of the evidence measurable.
Were Roman roads really straight?
Only when the ground made it possible. The study measured the sinuosity of one-kilometre road segments, comparing each segment’s actual length with the straight-line distance between its ends. Roman roads were generally quite direct, but their average sinuosity was higher than that of modern roads.
Flat areas such as parts of western Europe, the Po Valley, North Africa and the plains of Anatolia allowed straighter connections. Mountain roads were more complicated. In southeastern Europe and other rugged regions, routes bent around slopes, valleys and passes. The pattern is a reminder that Roman engineering did not erase geography; it negotiated with it.
Modern roads are usually straighter because machines can cut into hills, reshape ground and build for faster vehicles. A Roman road that appears to wander may therefore be showing the engineering limits of its time, rather than a lack of planning.
Cities were the hinges of the network
The study also asked how cities mediated movement. Provincial capitals were more centrally located in the road network than other cities, which fits their administrative role. They were places where people, goods and information could be redirected across a province.
Segments believed to have been major roads were also more important intermediaries than secondary routes. That sounds obvious until it is tested across an empire. The network model turns a label such as “main road” into a measurable question: how often did a route sit between places that needed to communicate?
Why modern routes still follow ancient corridors
One of the study’s clearest findings is a correlation between Roman road density and major modern roads. The result does not mean that every modern highway sits on Roman paving, or that the ancient engineers predicted the needs of cars. Valleys, river crossings, passes and settled plains still constrain movement today. The same geography can pull different societies toward the same corridor.
This is where the new research complements the surveying tools that helped Roman engineers set out their roads. Instruments and construction methods mattered, but so did the landscape they were trying to cross. A durable route could outlast the empire because it solved a problem that remained after the imperial government disappeared.
What network science adds to Roman history
Roads were more than military lines on a map. They structured the movement of food, trade goods, migrants, diseases, technologies and beliefs. A high-resolution network lets historians ask where those flows would have been easier, where a detour would have slowed them and which cities acted as unavoidable intermediaries.
The researchers describe their work as exploratory and descriptive. It does not by itself prove that a road caused a city to grow, or that the Roman network directly created a modern settlement. Those are harder causal questions. But a measured map gives future studies a firmer starting point than a collection of famous roads chosen one region at a time.
What remains unknown?
The map cannot restore roads that have vanished without a trace, and its coverage is not equally detailed everywhere. It also cannot tell us which traveller used each route or how seasonal weather changed the experience of a journey. Those questions still require archaeology, inscriptions, environmental evidence and local history.
Even with those limits, the study changes the mental picture. Roman roads were neither a perfectly straight grid nor a random set of tracks. They were a negotiated network: direct where the ground allowed it, adaptive where it did not, and persistent enough that modern routes often recognise the same opportunities in the landscape.
Primary source: Network science reveals the structure and lasting impact of the Roman road system, published in Nature Communications on September 15, 2026.
