Skip to content

The Gothic War of 376-382

Adrianople

An eastern Roman field army was destroyed and an emperor killed by a Gothic force outside Adrianople, in a defeat a contemporary historian called the heaviest since Cannae.

378 CE · Near Edirne, Turkey

The forces, and what the figures are worth

The eastern Roman empire

The emperor Valens

Evidence level: Disputed

Roman troops

Commonly estimated at 15,000 to 30,000

Modern reconstruction; Ammianus gives no total

Ammianus, the contemporary source, declines to give numbers. Estimates rest on late-Roman establishment strengths, which are themselves debated.

Evidence level: Probable

Roman dead

Two-thirds of the army

Ammianus Marcellinus 31.13

A proportion rather than a count, from a contemporary military man. Ammianus explicitly compares the scale to Cannae.

The Gothic coalition

Fritigern · Alatheus and Saphrax, commanding the cavalry

Evidence level: Unknown

Gothic warriors

No reliable figure survives

Not given

Ammianus gives no total. Estimates in the modern literature are conjectural.

What happened

Gothic groups admitted across the Danube in 376 under agreement were mistreated and exploited by Roman officials, and revolted. Two years of campaigning followed.

Valens advanced from Adrianople against Fritigern's wagon laager without waiting for the western army under Gratian, which was days away. Ammianus attributes the decision to a wish not to share the credit.

The Roman troops arrived after a long march in summer heat, without food, and were held in position while negotiations were attempted. Fires lit by the Goths added smoke and heat.

Fighting began before the negotiations concluded, and the absent Gothic cavalry returned and struck the Roman flank. The Roman line was compressed until, in Ammianus's description, men could not raise their weapons or draw their swords.

Tactics

The failures were operational rather than tactical: attacking without the converging force, committing exhausted troops without reconnaissance of the enemy's absent cavalry, and beginning an engagement during a parley. The compression of the Roman infantry into a mass too dense to fight is the same mechanism recorded at Cannae, and Ammianus's comparison is apt.

Consequence

The eastern field army had to be rebuilt. Theodosius settled the Goths inside the empire in 382 on terms allowing them to serve under their own leaders — a precedent with long consequences. The battle is often described as the moment cavalry displaced infantry as the dominant arm; that reading is not supported by the evidence and is treated below as a dispute.

Location

Evidence level: Probable

Outside Adrianople in Thrace

Modern: Near Edirne, Turkey

The city is certain; the battlefield has not been located.

Open questions

Evidence level: Disputed

Did Adrianople mark the end of infantry dominance?

A view associated with older military history, notably Charles Oman, held that Gothic cavalry inaugurated a thousand years of mounted supremacy. Modern scholarship rejects it: the Gothic force fought largely on foot, late Roman armies remained infantry-based for centuries, and the defeat is better explained by operational error than by a change in the balance of arms.

Primary sources

  • Ammianus Marcellinus, Res Gestae 31.12-13

    A contemporary account by a former staff officer, and one of the best-informed battle narratives in ancient literature.

How to read the evidence labels

Evidence level: Documented
Supported by a named primary text or by excavated material evidence.
Evidence level: Probable
Supported by strong inference that most specialists accept.
Evidence level: Disputed
Substantial specialist disagreement exists; competing readings are stated.
Evidence level: Literary
Belongs to poetic or mythological construction rather than to history.
Evidence level: Mythological
Belongs to shared religious or mythic tradition rather than to any single text.
Evidence level: Unknown
The available evidence is insufficient to decide.

These labels describe the status of the evidence, not our confidence in a conclusion. A claim marked Literary is not a claim we doubt; it is a claim about what kind of thing a poem is.