How Much of Romance of the Three Kingdoms Is True? Seven Parts History, Three Parts Fiction

The war is real. Cao Cao, Liu Bei, Sun Quan, Zhuge Liang and Sima Yi all lived, fought each other, and are documented in a court history compiled within living memory of the events. The novel that made them famous was written more than a thousand years later, by a man who had a side, working from a record that is still in print. Sorting one from the other is the single most searched question about this book, so here is the sorting.

• ✅The period is 220 to 280 AD, and the three states were founded in 220, 221 and 229 by three men who each declared himself emperor.
• ✅The history came first. Chen Shou finished his Records in the 280s or 290s, roughly a decade after the war ended.
• ✅The novel is 14th century, more than a thousand years after the fact, and the text everyone reads was re-edited again in 1679.
• ✅The most famous trick in it is the one piece of the Red Cliffs campaign the standard histories do not record.

Below, what the record actually says, what the novel added, and why the added version is the one the culture kept. If you would rather test a stratagem than read about one, the playable Romance of the Three Kingdoms will hand you the fog, the boats and three days to produce a hundred thousand arrows.

⚖️ The short answer, and who first gave it

Seven parts fact, three parts fiction is the formula, and it belongs to Zhang Xuecheng, a historian of the Qing dynasty who was assessing the novel as a historian would. It is worth attributing, because the ratio gets quoted constantly with no source attached, as if it were a measurement somebody took.

The seven parts are load-bearing and specific. Three states really did exist between 220 and 280 AD. Cao Cao's son Cao Pi took the throne of Wei in 220 when the last Han emperor abdicated. Liu Bei answered by proclaiming himself Emperor of Han on 15 May 221, and Sun Quan declared himself Emperor of Wu on 23 June 229. Every battle the novel is famous for was fought, in the year the novel says, by the commanders it names.

The three parts are where the pleasure is. Almost every scene a first-time reader remembers, the duels at the pass, the zither on the empty wall, the arrows in the fog, the oath in the orchard, sits in the invented fraction or has been reshaped past recognition. That is no complaint against the book; the invented fraction is precisely how it became the book everyone now argues about.

'Borrowing Arrows with Straw Boats' gives you the novel's version of the problem: three days, no bowyers, and a rival who set the deadline hoping you would miss it. Take Zhuge Liang's way out, or find your own. Role-play Romance of the Three Kingdoms →

📜 Chen Shou's Records against Luo Guanzhong's Romance

The two books are a thousand years and one enormous difference in purpose apart. Chen Shou was a court historian under the Jin dynasty, the regime that ended the war, and he compiled the Records of the Three Kingdoms in the 280s or 290s: sixty-five volumes divided into the Book of Wei, the Book of Shu and the Book of Wu, about 360,000 Chinese characters, organised as biographies rather than as a story.

Pei Songzhi is the figure worth knowing, because he is the reason we can check any of this. A century and a half after Chen Shou, in 429, he went back through the Records and attached every competing version of every episode he could find, quoting works that have since been lost entirely and saying plainly where accounts disagreed. His annotations are the archive the novelist later mined and the archive by which the novelist can be caught.

🏹 The arrows in the fog, and where they came from

The Battle of Red Cliffs happened in the winter of 208 AD, and it happened roughly the way the novel says: Cao Cao's enormous northern fleet, a southern alliance that should not have held, a captain called Huang Gai who offered a false surrender and sailed vessels packed with kindling, dry reeds and fatty oil into the middle of it. The fire ships are history; the arrows are the part of the campaign the record does not contain.

Chapter forty-six of the novel, which sits immediately before the four chapters the battle itself occupies, is one of the great set pieces in Chinese fiction. Zhou Yu, who wants Zhuge Liang dead and cannot say so, orders him to produce a hundred thousand arrows. Zhuge Liang volunteers to do it in three days instead of ten. He borrows twenty light boats, thirty men to a boat, screens them with blue cloth and lashes more than a thousand bundles of straw along each side. On the third night the river fogs over, he ropes the boats together and sails at Cao Cao's camp with the drums going. Ten thousand archers shoot into fog at a noise. He turns the line around so the arrows fill both sides, has his crews call out their thanks to the prime minister for the arrows, and comes home with five or six thousand shafts in every hull.

Standard accounts of the actual battle contain the false surrender and the fire ships and no arrow-borrowing at all. Zhuge Liang appears in them as the diplomat Liu Bei sent to talk Sun Quan into the alliance, which is a real and difficult job and not a stratagem on the water.

There is a real arrow story, though, and it belongs to somebody else. The Weilüe, a private history of Wei preserved inside Pei Songzhi's commentary, records that at Ruxu in 213 Sun Quan sailed out to look at Cao Cao's army, was shot at until his boat had taken so many arrows on one side that it began to list, and ordered it turned about so the other side could take arrows and bring it level. Chinese reference works treat that as the prototype of the novel's episode. A real incident involving the wrong man, in the wrong year, at the wrong battle, has been moved five years and handed to the character the novelist preferred.

🏆 Who actually won the Three Kingdoms?

Nobody the reader has been following. The dynasty that ends up holding all three states is Jin, founded by Sima Yan, the grandson of Sima Yi, who seized the throne in 266 and completed the conquest of Wu in 280 when Jianye fell in the third month and Sun Hao surrendered. Wei, Shu and Wu are all gone by then, and so is every character whose name you learned.

The novel does not soften this. Chapter one hundred and twenty says flatly that from this point the three kingdoms passed to the Jin emperor Sima Yan and became the foundation of a single realm, then quotes back the line about long division and long unity that opened the book. Sun Hao, the last emperor of Wu, is given a marquisate and a comfortable retirement. The last rulers of all three states die in their beds. The final couplet says the tripod of three legs has become a dream, and that those who come after can only mourn it for nothing.

This is the most-asked question about the book on English-language search, and it is asked because the answer feels wrong. Sixty years of loyalty, genius and sacrifice, and the prize goes to the family of the man who spent his career being outmanoeuvred by Zhuge Liang. That is also, as it happens, what the historical record says.

🖋️ Why the invented version won anyway

Because it was edited to. Every modern translation renders the revision Mao Lun and his son Mao Zonggang published in 1679 rather than the fourteenth-century text, and that revision was successful enough to drive its rivals off the market and remain, for close to three centuries, the only version most Chinese readers ever saw. The Maos did not merely tidy. Their edits quietly realign who the reader is supposed to be rooting for, and Mao Lun was unequivocal in praising Liu Bei and condemning Cao Cao where earlier texts had been more ambivalent.

That is why the arrows migrate to Zhuge Liang and why the empty city on a wall in chapter ninety-five is remembered as the cleverest thing anyone ever did, a scene taken apart in full here. A story with a favourite gives its favourite the best material. The same editorial thumb is what turned the most capable politician of the era into the villain of Chinese popular memory, which is its own argument.

None of this makes the novel dishonest. It makes it a novel. The useful habit for a first reader is simply to notice which scenes are doing emotional work and check those first, because the invented fraction and the memorable fraction turn out to be very nearly the same set.

Chapter numbers here were checked against the Chinese text at Project Gutenberg. If you are deciding which English edition to buy first, the translation comparison is here.