Roman roads 2

Roman road. Credit: Queen Mary University of London

Byzantium, Not Rome, Was the Empire’s Real Overland Hub, Map Shows

In A Nutshell

  • A new digital map of nearly 186,000 miles of Roman roads shows Rome wasn’t the empire’s top hub for overland travel.
  • Byzantium, Antioch, Ancyra and Corinth all ranked more central than Rome for moving people and goods across the empire by land.
  • Roman roads ran straight where the terrain allowed but were noticeably less straight overall than today’s major highways.
  • Roman road density still lines up with where major modern roads run today, especially in North Africa and the Near East.

Antioch, Ancyra, Byzantium and Corinth all rank ahead of Rome as modeled hubs for overland travel across the Roman Empire, according to a new digital map of nearly 186,000 miles of ancient roads, published in Nature Communications. For centuries, the phrase “all roads lead to Rome” has stood in for the idea that the capital pulled every route toward it. Researchers spent years digitizing the empire’s road network to test that claim, and their model of overland connectivity tells a different story.

Rome sat comfortably in the middle of the Italian peninsula, but in the model, that peninsula worked more like a side street compared with the busiest corridors running through North Africa, the Levant and the Mediterranean coast. Twenty-four of the empire’s 45 provincial capitals sat along roads ranked in the top fifth for empire-wide overland connectivity, while Rome itself was not located on the single most important road segment within Italy.

Built from the most detailed digital model yet of the Roman road system, spanning territory from Britain to the Middle East and North Africa, the analysis leans on network science, the same math used to study airline routes or social media connections, rather than assumptions passed down through decades of scholarship.

A New Dataset Maps Nearly 186,000 Miles of Roman Roads

Studying an empire that covered roughly 2.1 million square miles around the Mediterranean is no small task. Researchers pulled together a dataset tracking nearly 186,000 miles of roads, sorting them by confidence in their exact path, by whether they counted as major or secondary routes, and by how they interacted with mountains, rivers and valleys.

To figure out which places mattered most, the team measured betweenness: how often a stretch of road falls along the quickest modeled route between other points in the network, similar to how a busy highway interchange sees more traffic than a quiet side road. Each segment was weighted using travel-time estimates based on walking speed and slope, following a formula that accounts for how much slower travel gets uphill.

rome roads infographic
Nearly 186,000 miles of mapped Roman roads challenge the legend that all roads led to Rome. (Image by StudyFinds)

The Roman Road Network Didn’t Revolve Around Rome

Byzantium, now Istanbul, stood out in particular because it linked the European and Asian sides of the road network through ferry crossings at the Bosporus and the Dardanelles. Antioch, Ancyra and Corinth also outranked Rome, each occupying a more central spot on the network than the Italian capital.

Land travel is all this study measured. Water transport, both river and sea, mattered enormously for the Roman world, and Rome was extremely well positioned for maritime trade across the Mediterranean, so the finding does not knock Rome out of importance altogether. It simply shows other cities carried more of the empire-wide overland connectivity than the capital.

Roman Roads Weren’t as Straight as Legend Claims

Researchers also tested a second assumption: that Roman roads were famously, almost obsessively straight. The data tell a mixed story. Roads ran straight in flat regions like northwestern Europe, where there was little reason to curve around obstacles, but across the empire as a whole, they were noticeably less straight than today’s major highways, especially in mountainous terrain that forced detours and switchbacks.

Roads through the Alps had gentler slopes than roads through the Pyrenees or the Taurus Mountains, a pattern suggesting Roman builders worked hard to keep connections between Italy and the rest of Europe manageable, even when that meant cutting through rock.

Main roads, long assumed to be the empire’s major arteries because they show up in ancient itineraries and are marked by milestones, also proved their worth: across nearly all provinces studied, they scored higher on the travel-importance measure than secondary roads, giving numerical backing to what had mostly been an educated guess.

Roman Roads Still Shape Where Modern Roads Run

Comparing the ancient road data with a modern global roads dataset, researchers found a moderate positive relationship between where Roman roads were dense and where major modern roads are dense now, though the strength of that link varies widely by region. Some of the clearest local connections turned up in North Africa and the Near East, pushing back on theories that those regions lost their link to ancient infrastructure once wheeled transport declined there in the medieval period.

Areas with the biggest jump in road density since Roman times cluster around modern cities including Paris, Madrid and parts of Britain, places that grew from smaller Roman-era settlements into major metropolitan hubs. Greece, Asia Minor and North Africa, once densely urbanized under Roman rule, show comparatively lower modern road density today.

Empires build their own mythology, and “all roads lead to Rome” has survived two thousand years as a tidy summary of Roman ambition. Road-network patterns tell a messier story: an empire held together not by one dominant overland hub but by a handful of well-placed cities carrying much of the connective load. Rome remained the empire’s political capital and a major Mediterranean center by sea, but when researchers modeled overland roads alone, other cities held more central positions. For overland travel, the map gives places such as Byzantium and Antioch a stronger claim than the old saying suggests.


Paper Notes

Limitations

Researchers were upfront that their road dataset, while the most detailed built so far, is not complete. Coverage and reliability vary by region depending on how much archaeological survey and excavation work has happened there, and some areas, including central Europe, parts of the Balkans and several desert regions, have lower confidence data than well-studied regions like Italy, Greece or the Near East. The study also treats the road network as a single snapshot, largely representing the structure of the empire during the second through fourth centuries CE, even though the real network developed gradually and changed with military, political and economic circumstances. Future work, researchers noted, should track how the network changed over time rather than treating it as fixed. Their analysis also covers land travel only, not river or sea routes, which mattered a lot for long-distance movement in the Roman world.

Funding and Disclosures

Support for this research came from Danmarks Frie Forskningsfond (DFF) through a Sapere Aude research leadership grant for a project called MINERVA, a Carlsberg Foundation Semper Ardens Accelerate Fellowship for a project called The Past Social Networks Project, the Danish National Research Foundation’s Centre of Excellence for Urban Network Evolutions (UrbNet), and a Viator-e project funded by Spain’s MCIN/AEI and FEDER program. The authors reported no competing interests.

Publication Details

Titled “Network science reveals the structure and lasting impact of the Roman road system,” the study was published in Nature Communications in 2026. Authors include Adam Pažout, Tom Brughmans, Pau de Soto, Matteo Mazzamurro, María Coto-Sarmiento, Clara Filet, Eduardo Herrera Malatesta, Magnus L. Nielsen, Gustav Emil Ølgaard and Peter B. Vahlstrup, affiliated with institutions including Aarhus University, Universitat Autònoma de Barcelona, Queen Mary University of London, Milà i Fontanals Institution for Research and Humanities (IMF-CSIC), University of Bordeaux-Montaigne and Leiden University. The paper’s DOI is 10.1038/s41467-026-75453-3.

About StudyFinds Analysis

Called "brilliant," "fantastic," and "spot on" by scientists and researchers, our acclaimed StudyFinds Analysis articles are created using an exclusive AI-based model with complete human oversight by the StudyFinds Editorial Team. For these articles, we use an unparalleled LLM process across multiple systems to analyze entire journal papers, extract data, and create accurate, accessible content. Our writing and editing team proofreads and polishes each and every article before publishing. With recent studies showing that artificial intelligence can interpret scientific research as well as (or even better) than field experts and specialists, StudyFinds was among the earliest to adopt and test this technology before approving its widespread use on our site. We stand by our practice and continuously update our processes to ensure the very highest level of accuracy. Read our AI Policy (link below) for more information.

Our Editorial Process

StudyFinds publishes digestible, agenda-free, transparent research summaries that are intended to inform the reader as well as stir civil, educated debate. We do not agree nor disagree with any of the studies we post, rather, we encourage our readers to debate the veracity of the findings themselves. All articles published on StudyFinds are vetted by our editors prior to publication and include links back to the source or corresponding journal article, if possible.

Our Editorial Team

Steve Fink

Editor-in-Chief

John Anderer

Associate Editor