In 1683 at the Battle of Vienna, 20,000 Polish, German and Austrian cavalry charged the Ottoman lines in what Wikipedia says is the largest cavalry charge in history. Other references are more cautious: this page and one of the Wiki references say “largest known” while 10 Heroic Cavalry Charges says “one of the largest”.

There may be some European / American bias, though. Looking at the list of Notable Charges, almost all those listed are European or from the American Civil War. No mention of the Mongols or any of the armies of Chinese dynasties such as the Sui or the Ming. Of course, the American civil war was comparatively recent so figures are more likely to be available and reliable.

Nonetheless, considering the size of some of the armies that these eastern powers had, these omissions seem to be rather glaring. In Richard A. Gabriel’s Genghis Khan's Greatest General: Subotai the Valiant the author, referring to the Mongols, states that

The army was almost entirely composed of cavalry, with 40 percent heavy cavalry, and the remaining 60 percent designated light cavalry. There were no organic Mongol infantry units...

Gabriel also says that

The charge of the heavy cavalry was always the main player in the endgame of a Mongol attack.

As Genghis Khan could field an army of up to 100,000 Mongols (see also the Battle of Yehuling) it seems likely that some of the cavalry charges could easily have exceeded 20,000. Then there is the army of Fu Jian in 383 AD which apparently included 270,000 cavalry, and that of Yang Liang in 586 AD numbering “300,000, mostly composed of archer cavalry.” for the invasion of Korea.

There is, of course, a problem knowing the real numbers in many cases. For example, numbers for the Battle of Mohi in 1241 vary considerably.

I’ve been unable to find much information on the size of cavalry charges for either Chinese or Mongol armies and whether or not any of them surpassed in numbers the one at the Battle of Vienna in 1683.

Are there any primary sources on the size of cavalry charges for the Mongol or Chinese armies which are considered reliable by modern historians?

If not, have modern historians been able to make any estimates based on what evidence is available? Do we have any idea what proportion of the cavalry any of these armies might have committed to a single cavalry charge?

I would accept an answer for either the Chinese or the Mongols. Although I'm not asking about other armies (including Arab, Persian etc. might make the question too broad), I would be interested in reliable primary sources / estimates made which match or surpass the Vienna 1683 cavalry charge.

  • 9
    The field of battle limits the size of a cavalry charge; the horses of a single line can only be so close to each other, and typically only two lines of cavalry can charge at one time. These larger armies provide for (1) multiple engagements, (2) large reserves, (3) outriders to raid and protect their flanks. Another limitation is the ability to provide feed for the horses - you can hardly provide forage for 100,000 horses at a single site, so the cavalry units must be staged and moving. Commented Jun 15, 2018 at 18:29
  • 3
    @PeterDiehr Your points well made, but the Mongols fought some of their battles on the steppes (for example, against the Jin I think). Also, many of their commanders were tactically astute and would have sought to maximize battlefield advantage by choosing sites that were ideal for them (though this wouldn't always have been possible of course). Commented Jun 16, 2018 at 3:51
  • 2
    How much did the Mongols have in terms of cavalry which would charge? Weren't their tactics heavily based around missile cavelry?
    – user31561
    Commented Jun 19, 2018 at 18:55
  • 1
    What about the Maratha cavalry armies? They could feature as well...
    – gktscrk
    Commented May 6, 2020 at 10:44
  • 1
    @LarsBosteen: A very cursory glance thus far has found the Battle of Kars, 1745, where Nader Shah supposedly led 40,000 cavalry against the Ottoman flank and collapsed it. Does that suit? I'll dig some more though.
    – gktscrk
    Commented May 6, 2020 at 14:14

4 Answers 4


I suggested in the comments above, earlier, that the Hindustan peninsula might have some appropriate examples. A cursory look at the Moghul, Maratha and Bengali armies, etc, did show that the cavalry forces put into field were of the order of tens of thousands of cavalrymen (Second Panipat, Tukaroi), I did not find an immediate example of a combined charge.

The Battle of Kars, 1745

However, I found a possible contender during the Ottoman-Persian conflict of 1743-1746. In 1745, Nader Shah was conducting forward defense of his borders, and met the Ottomans at the Battle of Kars, 1745. In this, Nader Shah is reputed to have won the day by organising a flank attack with 40,000 of his finest cavalry at the right moment.

I will describe the specifics of this account below, along with some sources.

Vatatzes' Account

From Michael Axworthy's 'The Sword of Persia' (location 5070):

The armies closed and engaged, but the cavalry of both sides held back. The infantry struggle went on for some time, each side attacking and counter-attacking; both commanders sending new reinforcements into the battle. One account says that, contrary to his usual practice, Nader in this campaign against the Ottomans had generally commanded the army from his throne within the camp, receiving word of developments and issuing commands to his troops by messengers. But by the early afternoon on this day his messengers were still telling him the fight was undecided, and he resolved to intervene personally. Nader put on his armour, mounted up and led a reserve of 40,000 Abdali cavalry in a furious attack against the Ottoman flank. Even then the fighting was still fierce, and two of Nader's horses were killed under him, but his presence encouraged the Persians to new efforts. Ottoman resistance finally began to crumble, and 15,000 irregular troops from the provinces of Asia Minor fled. The main Ottoman body retreated in confusion to the safety of the entrenchments around their camp. Nader followed them up, but withdrew with his men to his own encampments at sunset.

An accompanying footnote clarifies this is primarily based on the account of Basile Vatatzes, a Greek traveller and merchant in Persia. Axworthy describes Vatatzes' 'additional detail' on the course of the battle as:

Vatatzes, whose account (pp. 282-3) gives colour to what was previously known of this battle, is the source that says the cavalry of both sides were initially kept back. He says the Persian cavalry were superior and the Ottoman cavalry feared to attack them (one of the sources presented by Sha'bani also says the Ottomans did not have the will for a cavalry fight - Sha'bani 1977, pp. 34-7). By contrast, Ottoman sources suggest that their troops came close to beating the Persians (Cambridge History of Iran, Vol. 7, p. 309).

I should note that I did not access Vatatzes' account myself.

Quality of Vatatzes' Account

A separate article by Axworthy, 'Basile Vatatzes and His History of NĀDER ŠĀH' describes Vatatzes's work and its merits:

is patchy, often inaccurate, and his judgement is often suspect. ... Frequently, he reports the mood and general drift of events better than the details.

Nevertheless, the examples brought in the paper do not specify any numeric errors (although chronological errors are brought up), and Axworthy specifically says that Vatatzes' military descriptions are particularly useful on the count of their accuracy (e.g., on Persian titles) and thoroughness (e.g., on drilling the troops).

Astarabadi's Account

An alternative source is Mirza Mehdi Khan Astarabadi whose history of Nader Shah has been accessible in French since 1770 as Histoire de Nader Chah. This book contains the following description of the battle of Kars, much simpler and almost no (useful detail):

Le dix, dans l'après-midi, Mohammed Pacha s'avança avec cent mille hommes de cava lerie, & quarante mille d'infanterie, & campa au pied d'une montagne à deux parasanges de l'armée impériale, où, ayant dressé ses tentes, il commença de fortifier les endroits foibles, & de préparer ses canons & ses mortiers.

L'onzième, les deux armées étant rangees en ordre de bataille, le feu du combat com mença, à flamber, & ses étincelles atteignirent les étoiles. Après plusieurs successifs engagemens, l'armée Ottomane fut mise en déroute par l'interposuion de la' Providence.

La perte sut très-grande du côté des Turcs, leur genéral se retira dans ses retranchemens, & la nuit devenant obscure, les troupes conquérantes retournèrent à leur camp.

An approximate (though good enough in my opinion) (Google) translation reads:

On the tenth, in the afternoon, Mohammed Pasha advanced with a hundred thousand cavalry men, & forty thousand infantry, & encamped at the foot of a mountain with two parasanges of the imperial army, where, having erected his tents, he began to fortify the weak places, and to prepare his guns and his mortars.

The eleventh, the two armies being placed in order of battle, the fire of battle began to flame, and its sparks reached the stars. After several successive engagements, the Ottoman army was routed by the interposition of Providence.

The loss was very great on the side of the Turks, their general withdrew into his retrenchments, and at night becoming dark, the conquering troops returned to their camp.

Quality of Astarabadi's Account

The quality and level of detail is clearly without the colour that Vatatzes provided for Axworthy's account. Nevertheless, the quality of this writer is also in doubt (again by Axworthy, from "The Sword of Persia"):

The Jones translation contains many errors, particularly with dates and names, but is an accessible version for those who cannot read Persian... Mirza Mahdi was Nader's official historian, and understandably avoided including anything critical of Nader in his work. There are also indications that by the time the work was finished (about a decade after Nader's death), Mirza Mahdi needed to avoid offending some of the leading personalities of the Qajar tribe.

Nader Shah's Cavalry

I suppose to conclude, I should sum up what I noted on the Persian cavalry to support the argument that they could have done this (though, indeed, there are many cavalry units which could have done such a charge but did not).

Specifically, in another article on the Persian Army of Nader, Axworthy notes the particular quality of the Afghan (Abdali / Durrani) cavalry:

...his Afghan cavalry were probably the finest shock cavalry in the region.

Though the above article was illuminating in a general sense, and I'd probably recommend a read of it if the above topic sounds interesting, it doesn't bear much more discussion except to note that Nader Shah had plenty of troops armed and armoured exactly to fit the heavy shock trooper role as expected from a cavalry charge, and the troops most suitable are the ones that Vatatzes has placed into the role at Kars.

This is, regrettably, slightly circular as Axworthy's main source on the army is Vatatzes, of course, so this caveat should be borne in mind.

The numbers do, however, bear up for the general proportions of the army, with Axworthy quoting Jonas Hanway for 1744 (excluding some other larger tribes):

There were 13,000 guard cavalry, 20,000 cavalry from Nader’s own Afshar tribe, 50,000 Afghan cavalry, 12,000 jazayerchis and 40,000 ordinary musketeers. There were also 18,000 Turkmen, Uzbeks, and Baluchis, who served as light troops.

It should be noted that Nader Shah's son was conducting a separate campaign at the same time, winning a simultaneous battle near Mosul which completely destroyed the Turkish capacity for defense. Therefore, a good proportion of troops may have been with the son, making this charge unlikely, though Astarabadi implies that these were not troops from the main army (p 100):

Sur cet avis fa Majesté envoya le prince Nasralla Mirza pour s'opposer à ceux qui s'approchoient des frontières de Perse, & lui donna les légions victorieuses qui avoient été employées sur les confins de Karmanchah, du Loristan, & du Kiurdestan.

Same Google Translate option:

On this advice the Majesty sent Prince Nasralla Mirza to oppose those who approached the frontiers of Persia, and gave him the victorious legions which had been employed on the borders of Karmanchah, Loristan, & Kiurdestan.

Overall, on the balance of probabilities, I think that this here is a larger cavalry charge than at Vienna 1683, even if it does not stretch to the exact 40,000 as quoted by Axworthy, following from Vatatzes. Nevertheless, I think that if the numbers were ridiculous or the action improbable, Axworthy would have noted this much as he criticised Vatatzes for other reasons. Therefore, I would suggest that the largest cavalry charge we know of took place during the Battle of Kars.

  • 3
    Good job on sourcing and research. Even if 40,000 is an exaggeration, even half that number would match Vienna. I'll just wait a day or two before accepting to see if anyone else wants to have a go at this. Thanks for your effort! Commented May 6, 2020 at 22:58
  • 1
    @LarsBosteen: Thanks; much appreciated. From what it looked like, understanding of Arabic, Sanskrit, etc, is necessary for good sourcing of most of the other battles in the area. I should also say that my first thought was actually the Sengoku Jidai, and I saw that Takeda Shingen is known to have organized a charge of approx 10,000 cavalry, supported by infantry. Given he was only one of many daimyo there could be more to follow up on there, perhaps during the Korean invasion, but Central Asia was a more promising line of thought (as evidenced above).
    – gktscrk
    Commented May 7, 2020 at 4:34

While both Mongolian and Chinese mounted armies could easily outnumber those of the American or European at any point in history, it should be noted that the Mongols or Chinese hosts didn't really rely on cavalry shock charges as the Europeans or Americans did. Chinese and Mongolian cavalry were as you mentioned horse archers and wouldn't be used for the same purpose as European shock cavalry. The Polish Winged Hussars used by John Sobieski in the Battle of Vienna were much heavier and meant to cause fear in the enemy lines as well as to route enemy troops. In comparison, the Mongol and Chinese horse archers were meant to be lighter equipped and used for harassment or hit and run tactics to defeat their often unmounted or more heavily mounted foes.

In summary, yes the Chinese and Mongols definitely had much larger hosts of mounted units, but the Europeans win in crazy all out charges towards the enemies front lines

  • 6
    Sources would improve this answer. Also, while it is true that most of the Mongolian cavalry consisted of horse archers (as stated in the question), there was a substantial heavy cavalry element too. Commented Jul 8, 2018 at 22:50

Were Mongol or Chinese cavalry charges among the largest in history?

The Mongol's effectiveness in battle was not based on "riding over" their opponents; nor was it based on having superior numbers. It was based upon being the only side which both possessed the super weapon of the day, the Mongol Bow, and enough talented archer/horsemen who could wield it.

As such the Mongols were less likely to involve themselves in an early cavalry charge and were much more likely to use the mobility of their horses to keep them just out of range of enemy archers while their own mounted archers attrited their opponents.

Take the battle of Mohi for example in 1241. European Cavalry vastly outnumbered the Mongols. European forces numbered 80,000 Primarily light cavalry and the Mongols numbered roughly 30,000 Primarily horse mounted archers. The Hungarian cavalry was among the best in Europe at the time. The Mongol's victory was still lopsided 10,000 European casualties, Mongols suffered a few Hundred.

This lopsided victory is credited with Mongol tactics which took advantage of the superior range of it's horse mounted archers, and not because the Mongols fought like traditional European Cavalry and road over their opponents.

The largest Mongol / Chinese battle was the Battle of Yehuling which was a series of 4 battles 2 sieges and 2 segments fought in terrain chosen by the Chinese to negate the Mongol cavalry advantage (mountains). Huge Cavalry charges weren't present in any of these battles.

  • Battle of Wusha Fortress - was fought by the Chinese behind the Great Wall attempting to stop the Mongol invasion before it began.

  • Battle of Yehuling and Huan'erzui - were fought in mountainous regions chosen by the Chinese because they thought to make it harder for the Mongol cavalry to traverse up and down the mountains. It did, but it also made it harder for the Chinese to mass their superior number of troops.

  • Battle of Huihe Fortress - Another siege. The cavalry charge which decided the battle was personally led by Genghis Khan but only involved 3,000 mongols.

  • While not providing a bigger cavalry charge than the OP, you do answer the specific question better than I did (a "no" to the Mongols)! :)
    – gktscrk
    Commented May 21, 2020 at 18:11

The Mongol charges were not massed charges but independent hammer blows along a preferably strung out front. The aim was to achieve breakthroughs and threaten vital rear objectives. 10k to 15 k cavalry charges would have happened but no more. Their goal was never to allow the larger slower more logistics dependent armies to catch them out on the field in a static pitched battle but to create confusion and constantly keep them on the move. Then they would isolate select and systematically destroy elements as required.

Yehuling or badger mouth involved a series of three battles not one and they used tumens (10k cav in theory often lower in practice) in systematic strikes while remaining in contact with each other. So yes the case of entire mongol army massed as a single cavalry unit charging as a single block is highly unlikely.

  • 6
    Welcome to History:SE. Sources to support your assertions would greatly improve your answer. You might find it helpful to review our site tour and Help Centre and, in particular, How to write a good answer. Commented Mar 12, 2020 at 1:23
  • 2
    I actually love the info in this answer. However, Sem is right that we generally expect references on this site for non-trivial assertions of fact. Would it be possible to add some? I'd love to be able to upvote this.
    – T.E.D.
    Commented May 20, 2020 at 21:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.