The Persian Immortals: The Ten-Thousand Elite of the Achaemenid Empire
“This corps was known as the Immortals, because, whenever one of its members was killed or fell sick, another was at once chosen in his place, so that the number of ten thousand was never exceeded or reduced.”
Herodotus, Histories VII.83.
That is the unit, more or less, that we have been arguing about for two and a half millennia. Read carefully, Herodotus tells us several things at once: the athánatoi (ἀθάνατοι) — “the Undying” — were a closed-roster corps of exactly 10,000; they were all Persians; the immediate-replacement mechanism gave the unit its Greek name; and the figure of 10,000 was treated by Herodotus himself as exact. Almost everything else about the regiment is contested. The number, the recruitment, the equipment, the tactical doctrine, even the question of whether they were 10,000 or 1,000 — these are matters on which modern scholars disagree, sometimes sharply. I will come back to each. The reading here leans on the standard modern philology: Maria Brosius, The Persian Empire from Cyrus II to Artaxerxes I (2020); Pierre Briant, From Cyrus to Alexander (2002); Margaret Cool Root, The King and Kingship in Achaemenid Art (1979); and Matthew Stolper’s foundational work on the Persepolis Fortification Tablets, on which much of our hard administrative evidence rests.
A note on the title. The Greek athánatoi is itself a translation. The Persian form is usually reconstructed as something like anaušā (“immortal”), but the actual Old Persian word for the unit is not in the corpus, and the most we can say is that Herodotus, Ctesias, and Aelian, writing in three different centuries, are all translating the same title. The claim that the word is a “translation” of a Persian term is therefore a hypothesis, not a fact. I will use “Immortals” throughout as a transliteration, not as a translation.
The number, in detail
Let me put my cards on the table. I think the figure of 10,000 is approximately right, but is unlikely to have been exact. Three things support it: (1) the administrative scale of the Achaemenid state, which the Persepolis Fortification Tablets have made tangible since Hallock’s 1969 publication; (2) the cross-source consistency — Ctesias, Aelian, and later writers repeat Herodotus’ figure with minor variations; (3) the Achaemenid practice of organising forces in multiples of 1,000, the hazārā (a hazārpatish, a chiliarch, commanded each), which makes 10,000 a natural round number. The Achaemenid state had the demographic and organisational capacity to field a guard of that size, and the figure 10,000 is the right order of magnitude for an Achaemenid royal bodyguard of the sixth to fourth centuries BCE.
Where I think Herodotus is wrong is in the exactness. The replacement system, in real life, is unlikely to have been instantaneous. A trained archer-spearman, equipped in the standard Achaemenid panoply, takes months to train, and the regimental command cannot have absorbed a casualty in real time. The “perpetually 10,000” formula is a literary construction — Herodotus’ readers want a tidy figure, and Herodotus gives them one. The reality was, I suspect, closer to a target strength of 10,000 that was maintained within reasonable bounds (say 9,000–10,500) over time, with the under-strength treated as a regimental matter. The Sasanian royal guard, the pushtigbān of the late Sasanian period, was much smaller — a few thousand — and the institutional continuity between Achaemenid and Sasanian is a continuity of function, not of name or number.
Composition
The Immortals were not ethnic Persians in the narrow sense. The Achaemenid imperial guard, like the Achaemenid court generally, was a multi-ethnic body drawn from the senior Iranian peoples. The Persians and the Medes provided the core; the Sogdians, Bactrians, Parthians, and other eastern Iranian groups contributed contingents. The Macedonian sources — Arrian, Quintus Curtius, Plutarch — give the unit at Alexander’s invasion (334–330 BCE) as still 10,000-strong, still all Persians, still in the centre near the king. The Macedonian picture is the literary descendant of Herodotus, not an independent report. The Macedonian descriptions of the regiment at Gaugamela in 331 BCE are detailed enough to make the picture vivid: the Immortals in the centre, with the Persian and Median kinsmen of the king, the apple-butted spears conspicuous, the gold and silver ornaments of the regimental officers. Alexander’s decisive blow, on the Macedonian account, was the cavalry charge that opened the Persian line at the centre; the Immortals, caught in the disintegration, were broken rather than killed. The Macedonian picture is partly literary and partly based on the actual appearance of the unit.
Recruitment rested on a feudal obligation inherited from the Medes. Each noble family was bound to provide a fixed number of warriors for royal service; the avada system, in effect, was the labour service of the Achaemenid aristocracy. Service in the Immortals was the highest honour a Persian nobleman could be asked to give, and the equipment — gold ornaments, fine Median textiles, the long spear with the apple-shaped buttspike — marks the aristocratic character of the corps. The buttspike is the visual signature: the gold or silver apple at the butt of the spear was a status marker visible across the field.
Weapons and dress
The standard arms were the composite recurve bow, the long spear with apple-shaped buttspike, the short sword (akinakes in Greek, acinaces in the Herodotean spelling — both forms are used in classical scholarship), and the wicker shield. The bow was the practical weapon at distance; the spear was the close-combat weapon; the sword was the personal weapon. Helmets were of bronze or iron, of types influenced by Greek, Assyrian, and Median traditions. Body armour, in the Achaemenid period, was light by Greek standards; the Persians relied on mobility and the bow, not on the bronze plate of the hoplite.
Dress was elaborate. The Persian wore loose trousers, a long-sleeved tunic, a soft felt cap, and a cloak; the Median wore a long outer robe and a more ornate cap; eastern contingents wore local variations. The overall effect, to Greek eyes, was theatrical. Herodotus lingers on the gold ornaments, the bracelets, the decorated pommels, and the regiment on the march must have presented a striking sight. By the time of Xerxes I the regiment may have adopted a more uniform appearance, based on the court dress of Persepolis. The Persepolis reliefs show the dress evolving over the two centuries of the dynasty; by the late fourth century BCE the equipment is recognisably distinct from the early sixth.
Tactics
The Achaemenid infantry did not fight as a Greek-style shock force. The standard deployment — described by Herodotus, Xenophon, and the reliefs alike — was the mass of infantry in the centre, the cavalry on the wings, and the king with his guard held in central reserve. The Immortals were not a phalanx. They were a heavy multi-purpose force trained to fight in dense formation with spear and sword, to skirmish with the bow, to hold a critical point, to relieve a hard-pressed unit, and to exploit a breakthrough. This flexible doctrine, well adapted to a multi-ethnic army, is the distinctive feature of the Persian way of war.
The Immortals were held back from the heaviest fighting in the great battles of 480–479 BCE. At Thermopylae, Xerxes committed them at the decisive moment, and they failed to break the Greek line. The pass was taken by the flanking movement of Hydarnes through the Anopaea path; Ephialtes was the local guide, not the commander. At Salamis the Immortals had no role — the battle was a naval engagement. At Plataea in 479 BCE, Mardonius’ army was broken by the Greek hoplites under Pausanias, and the Immortals are mentioned as the central body of the Persian line. The standard reading — that they were killed almost to a man — is the natural inference from Herodotus, but it is not a certain one, and modern scholars hedge. The Macedonian sources, writing a century and a half later, give the regiment at 10,000 strong in the battles of Issus (333 BCE) and Gaugamela (331 BCE), which is improbable if Plataea had destroyed them.
What we don’t know
Two things would settle the question. One is a Middle Persian or Elamite administrative document listing the regiment by name and giving its strength; the Persepolis Fortification Tablets, in the published corpus, do not contain such a document, though new tablets are still being published. The other is a contemporary Achaemenid relief showing the regiment by name and size, with a casualty list; no such relief exists. Until either turns up, the question of “exactly how many” is not answerable in a strict sense. The order of magnitude — somewhere between 1,000 and 10,000 — is what the evidence will bear.
The wider context is in the Persian Military and Warfare survey, the Persian Cavalry cluster, and the How Many Persian Immortals Were There? long-tail page. The question of equipment is in What Weapons Did Ancient Persian Soldiers Use?. The Greco-Persian Wars are treated in How Did the Persians Fight the Greeks?.
The afterlife of the unit
The regiment, as a distinct unit, disappeared with the fall of the Achaemenid dynasty, but the institutional pattern — a closed-roster royal guard, multi-ethnic, with elite equipment and a central tactical role — was inherited by the Hellenistic successor states and adapted to the new political order. The Seleucid kings maintained a somatophylax, a personal bodyguard of fixed size, drawn in part from the Iranian noble families of the former Achaemenid court; the Ptolemaic kings in Egypt, who ruled over a large population of former Achaemenid subjects, maintained a parallel somatophylax of similar organisation. The tradition of a ten-thousand-strong royal guard, in Hellenised form, passed into the Seleucid and Ptolemaic courts, and the surviving members of the Achaemenid Immortals, after Gaugamela, were incorporated into the Macedonian army as the agema of the new Macedonian empire.
The Sasanian guard, sometimes referred to in later Islamic-period sources as the pushtigbān (literally, “those at the back”), was a smaller institution, a few thousand at most, and the continuity between the Achaemenid Immortals and the Sasanian guard is a continuity of function — the protection of the king’s person, the maintenance of royal dignity, the provision of a strategic reserve — rather than a continuity of name or size. The Sasanian guard was supplemented by the dihqān class of provincial nobility, who served the crown in war in return for land grants under the deh system.
The name in the modern world
The name has had a long afterlife in the modern world. The Immortals of the French revolutionary army, the elite units of nineteenth-century Latin American armies, the royal guards of the Qajar and Pahlavi dynasties, and the various “Immortals” units of modern Iran’s Islamic Revolutionary Guard Corps all borrow the name, and most borrow something of the idea. The Imperial Iranian Army under the Pahlavis maintained a Javidan (Immortal) infantry division as a ceremonial unit, and the modern Iranian military still uses the term in some regimental names. Frank Miller’s graphic novel 300 and the 2006 film introduced the regiment to a generation that had never heard of Herodotus, and the spired helmet, the gold pectoral, and the long spear tipped with an apple-shaped pommel are now firmly part of the iconography of the ancient world in popular culture.
The scholarly afterlife is no less striking: monographs by E. M. Yamauchi, Margaret Cool Root, and others have brought the Persepolis reliefs, the Greek literary sources, and the Babylonian and Egyptian administrative documents to bear on the question, and the result is a picture of an institution that was real, important, and substantially more complicated than the legend suggests. Maria Brosius’s The Persian Empire from Cyrus II to Artaxerxes I (2020) is the most recent comprehensive treatment in English, and the French and German scholarship on the Achaemenid military establishment, which has been particularly active since the 1990s, is gradually putting the unit on a firmer documentary footing.