Battle of Chilcheollyang

The Battle of Chilcheollyang was the greatest defeat of the Korean navy in the Imjin War. It came at the start of the second Japanese invasion, which began in 1597 after peace talks between Ming China and Japan had collapsed. The officer who first reported it to the court said that the attack began in the second watch of the night (roughly 9–11 p.m.) of the 15th day of the 7th lunar month and that the Korean fleet was destroyed the next morning. Converted with the table of the Korea Astronomy and Space Science Institute (KASI), those days are 27 and 28 August 1597.

The Joseon fleet was then commanded by Won Gyun. He had replaced Yi Sun-sin, the victor of Hansan Island, who had been arrested earlier that year for failing to carry out an order to intercept the Japanese commander Katō Kiyomasa at sea. Under heavy pressure from the court and the commander-in-chief Gwon Yul, Won Gyun led the fleet out toward Busan, where the Japanese were landing reinforcements. After days of exhausting movement and losses near Busan and Gadeok Island, the fleet fell back and anchored in Chilcheollyang, the narrow strait between Geoje Island and the island of Chilcheon.

There it was surprised at night by Japanese ships, and at dawn a much larger Japanese fleet surrounded it, while Japanese troops waited on land. Most of the Korean ships were burned or captured. The commanders Yi Eok-gi and Choe Ho died, and Won Gyun, who fled ashore, is reported to have been killed. Bae Seol, the naval commander of Gyeongsang Right Province, escaped with about twelve ships.

The defeat opened the south coast to the Japanese, who pushed west into Jeolla Province and took Namwon a month later. It also brought Yi Sun-sin back: the court reappointed him a week after the battle, and the few surviving ships became the core of the fleet he led at Myeongnyang in October 1597. The later sections look at how the disaster came about, what the sources say about losses and Won Gyun’s death, and how the blame has been argued over ever since.

During the long truce after 1593, Konishi Yukinaga, the Japanese commander most involved in the peace talks, kept in touch with the Korean army commander of Gyeongsang Right Province, Kim Eung-seo, through a go-between called Yoshira. Korean records describe him as one of Konishi’s men; the Encyclopedia of Korean Culture calls him a double agent. Konishi and Katō Kiyomasa were known rivals, and Yoshira’s messages suggested that Konishi wanted the talks to succeed and blamed Katō for their failure.

In the 1st lunar month of 1597 Kim Eung-seo reported that Yoshira had come to him on the 11th with a message from Konishi: Katō, with 7,000 men, had reached Tsushima on the 4th and would cross as soon as the wind allowed. Yoshira urged that the Korean navy move to Geoje Island and to the coast near Gijang to watch for Katō’s crossing, or at least show itself in force so that Katō would hesitate. The court took the information seriously, but the Border Defence Council also warned that “the words of the Japanese are cunning and hard to rely on” and left the decision to the commander-in-chief.

Yi Sun-sin did not take the fleet out to intercept Katō, and Katō landed on the south-east coast. According to the Revised Annals of King Seonjo, compiled in 1657, Yi judged that the sea route was dangerous and that the Japanese would surely lay an ambush: a large fleet would be seen, and a small one would be attacked. The Revised Annals add that Katō came past Dadaepo and went on to Seosaengpo, and Yu Seong-ryong’s Jingbirok says that Yoshira then came again, feigning regret that the Koreans had missed their chance.

Whether the whole affair was a Japanese ploy is not certain. The Revised Annals and Jingbirok present it as a trick by Konishi to lure out the Korean fleet and to bring about Yi’s downfall, and many modern Korean accounts repeat this. Other readings are possible: Konishi’s messages may also have been a move in his rivalry with Katō. The timing matters too: the warnings came while Katō was already on Tsushima waiting for a wind, and the Border Defence Council itself worried that it might be too late to act; the records used here do not show whether an interception at sea was practicable. What the records show clearly is how the message was used at court.

King Seonjo was furious. At an audience on the 27th day of the 1st month he said that Yi had done nothing since 1592 and had now let slip a chance that “Heaven had given”, and that even if Yi brought him Katō’s head he could not be forgiven. Some ministers defended Yi or warned against changing commanders in wartime; others pressed for his removal. The next day the king wrote to Won Gyun, whom he had just made naval commander of Gyeongsang Right Province, that Yi had “let the enemy cross the sea in peace” and must one day be arrested and interrogated.

In the 2nd lunar month the Office of the Inspector-General demanded Yi’s arrest, and the king ordered that he be arrested once Won Gyun had taken over his command, but not in the middle of a battle. Yi handed over the command at Hansan Island, his headquarters since 1593, and was taken to the capital; he was imprisoned on the 4th day of the 3rd month (19 April 1597). On the 13th the king listed his crimes: deceiving the court, letting the enemy go unpunished, stealing others’ merit and bringing false charges against others, and acting without any restraint; by law, he wrote, Yi should be put to death.

Yi was interrogated under torture. The elder statesman Jeong Tak argued that Yi was a famous general whom the state could not afford to kill and that commanders’ decisions could not be judged from afar. On the 1st day of the 4th month (16 May 1597) Yi was released and ordered to serve without rank (baegui jonggun) under the commander-in-chief Gwon Yul. On his way south he learned that his mother had died.

Won Gyun had fought beside Yi in 1592, and their quarrel over credit for those victories had grown bitter. He had often said that Yi failed to attack; now he was made Naval Commander of the Three Provinces and expected to do what Yi had not. The court records show that Won Gyun soon found the task as difficult as Yi had said it was.

By the middle of the 7th lunar month the Joseon fleet, worn out by a failed sortie toward Busan, had fallen back to the strait between Geoje Island and Chilcheon Island. Jingbirok says that Bae Seol warned Won Gyun that the waters off Chilcheon Island were shallow and narrow and that the fleet should move, but that Won Gyun refused. The Japanese, who held forts along this coast, could watch the Korean ships from the shore; Jingbirok says their camps passed word to one another as the fleet moved. According to the Encyclopedia of Korean Culture, the naval commanders Tōdō Takatora, Wakisaka Yasuharu and Katō Yoshiaki moved to the north of Geoje Island on the 14th and planned a combined attack from sea and land.

The first report to reach the court came from Kim Sik, a royal messenger who was with the fleet. He said that in the second watch of the night of the 15th, five or six Japanese ships suddenly attacked and set fire to four Korean warships. The Korean commanders hurriedly formed up. At cockcrow countless Japanese ships surrounded them three or four deep, filling the sea around the nearby islands. The Koreans fought while falling back, but could not hold and retreated to Chuwonpo in the Goseong area, where all the warships were burned and commanders and men were burned or drowned.

Yu Seong-ryong told the king a similar story: the Japanese crept in under cover of darkness, burned four ships, and the next morning surrounded the fleet; the Koreans were forced toward Goseong and went ashore, but Japanese troops had landed first and were waiting for them. Japanese accounts, as summarised in Japanese Wikipedia, say the Japanese fleet attacked from dawn on the 16th while troops on land supported it, and claim that thousands of Koreans were killed and more than 160 ships captured.

Kim Sik escaped ashore with Won Gyun and the magistrate of Suncheon, U Chi-jeok. He reported that Won Gyun, old and unable to keep up, sat down under a pine tree with only his sword, and that as he himself ran he looked back and saw six or seven Japanese with swords reaching Won Gyun; what happened next he could not say. Bae Seol and the commanders of Okpo and Angolpo got away, Kim Sik reported, and countless Japanese ships headed for Hansan Island.

The strategy that led to Chilcheollyang was set in Hanseong. The court wanted the navy to cut the sea route between Japan and Busan so that Japanese reinforcements could not cross. In the 6th lunar month the Border Defence Council proposed that the fleet be divided into squadrons under Bae Seol, Yi Eok-gi, Choe Ho and Won Gyun, which would take turns cruising the sea while Hansan remained the base, and complained that the navy was hiding in its harbours instead of showing itself.

Won Gyun disagreed. In a report of the 3rd month he argued that the Japanese at Angolpo and Gadeok Island were few and that the army should first drive them out by land; with “more than 300,000” picked soldiers, a great combined campaign by land and sea could decide the war. In the 6th month he again wrote that his fleet could not reach Busan safely while the Japanese held Angolpo, Gadeok and Jukdo, because it would have no anchorage ahead and enemies behind. The court replied that the decision belonged to the commander-in-chief and the chief state commissioner, who wanted the navy to sail.

The pressure rose. Gwon Yul reported in the 6th month that Won Gyun kept asking for the land attack first and had no intention of sailing out; Gwon had rebuked him repeatedly, and only when Nam I-gong, sent by the chief state commissioner Yi Won-ik, came to Hansan and sat there urging him on did the fleet sail. On the 10th day of the 7th month the king ordered Won Gyun to be told that if he again shrank back and let the enemy go, the law of the state would apply and he could not spare him.

The Revised Annals and Jingbirok both report that Gwon Yul had Won Gyun beaten, the Revised Annals because he feared the enemy and delayed, Jingbirok because he had achieved nothing. They also disagree about when: the Revised Annals put the beating before the final sortie toward Busan, Jingbirok after the fleet had retreated to Chilcheon Island, and the Encyclopedia of Korean Culture after a failed attack by Bae Seol near Ungcheon. A report from Gwon Yul that reached the court after the defeat shows that the three provincial naval commanders had told Won Gyun that they knew defeat was certain, that Busan and Jeolyeongdo could not be approached, and that a general in the field need not obey every order of his ruler; Gwon asked the court to punish them.

The fleet sailed from Hansan on the 18th day of the 6th lunar month (31 July) with about a hundred ships. The next day it attacked the Japanese at Angolpo and Gadeok Island. The report of Yi Won-ik’s aide Nam I-gong describes a hard fight in which Japanese ships were driven off and two captured; but An Hong-guk, the magistrate of Boseong, was killed by a bullet through the forehead, and the fleet returned to Hansan.

Early in the 7th month the fleet went out again. Yi Won-ik reported that the Right Navy moved across by night on the 7th and lay off Dadaepo near Busan, capturing ten-odd Japanese ships on the 8th (20 August), the day more than 600 Japanese ships arrived at Busan from Japan. The final sortie toward Jeolyeongdo (today’s Yeongdo, Busan) is described in later accounts, and its exact relationship to this dated operation is uncertain. Jingbirok says that the wind rose and night fell, and that the rowers, who had rowed all day from Hansan, were hungry, thirsty and exhausted. Both Jingbirok and the Revised Annals say that the Japanese ships came close and then slipped away without fighting, to tire the Koreans out, and that in the night the Korean ships were scattered by the strong wind.

Won Gyun gathered the remaining ships and made for Gadeok Island, where the thirsty men went ashore for water. Jingbirok says that Japanese troops burst out from inside the island and that the fleet lost more than 400 officers and men before retreating to Chilcheon Island. The Revised Annals also describe the ambush, but give no figure and move straight on to the retreat to Chuwonpo in Goseong. At court, an official later remarked that in the fighting on the 9th the soldiers had been so frightened that they could not shoot a single arrow. By the time it anchored at Chilcheollyang, the navy had lost men, ships and confidence, and was close to the Japanese-held coast.

Won Gyun (1540–1597) was the naval commander of Gyeongsang Right Province when the war began in 1592. His own fleet collapsed in the first days of the invasion, but with Yi Sun-sin’s help he took part in the victories of 1592. The two men fell out over credit for those victories, and after Yi became Naval Commander of the Three Provinces in 1593, Won Gyun was moved to army posts. In 1597 he returned to the navy in Yi’s place. What happened to him at Chilcheollyang is reported differently in the sources, as the advanced section on sources explains.

Yi Eok-gi, naval commander of Jeolla Right Province, had fought beside Yi Sun-sin since 1592, at Hansan Island among other battles; the Encyclopedia of Korean Culture says that he wrote to ministers defending Yi after his arrest. Choe Ho, naval commander of Chungcheong Province, had helped suppress the rebellion of Yi Mong-hak in 1596. Both died at Chilcheollyang; the Revised Annals say that they threw themselves into the sea, and Jingbirok that Yi Eok-gi leapt from his ship.

Bae Seol, naval commander of Gyeongsang Right Province, is the most controversial figure of the battle. Jingbirok says that he had warned Won Gyun repeatedly, secretly kept his own ships ready and fled when the Japanese attacked, so that his squadron alone remained intact; the Revised Annals put it more harshly, saying that his flight made the whole army collapse. He reached Hansan Island, burned the buildings, grain and weapons there so that the Japanese could not use them, and sent the remaining people away. The Encyclopedia of Korean Culture says that he left with twelve ships.

Bae Seol served under Yi Sun-sin for a time after Yi’s reappointment, but before the Battle of Myeongnyang he left the fleet, with permission to treat an illness, and did not return. He was caught at Seonsan in 1599 and executed in Seoul. The twelve or so ships he had saved stayed with the fleet and became the core of the navy Yi rebuilt.

Kim Sik’s report reached the court on the 22nd day of the 7th month (3 September 1597). The Veritable Records describe a long silence in which no minister could answer the king. King Seonjo said that the defeat was the will of Heaven, asked why the fleet had not set scouts or fallen back to Hansan, and remarked that Hideyoshi had always said the Korean navy must be destroyed first before the army could be taken. He also said that Won Gyun had not wanted to go at first and that the commander-in-chief’s pressure had caused the disaster.

On the same day Yi Sun-sin was reappointed Naval Commander of the Three Provinces. According to the Revised Annals and Jingbirok, the ministers Kim Myeong-won and Yi Hang-bok urged that only Yi could restore the navy. Gwon Yul had already sent Yi to the coast to gather the scattered ships; with very few men he travelled from Gyeongsang into Jeolla, collecting what was left. The ships Bae Seol had saved became his fleet.

With the Korean navy gone, the Japanese controlled the south coast. Jingbirok describes their advance west after the victory: Namhae and Suncheon fell one after another, Japanese ships landed troops at Duchijin, and the army besieged Namwon, which was held by the Ming general Yang Yuan with 3,000 men. King Seonjo had foreseen this danger on the day the defeat was reported. Namwon fell in the middle of the 8th lunar month (late September 1597), and its fall was followed by a massacre.

In Hanseong the king feared that the Japanese ships would now sail into the western sea and reach Chungcheong and Gyeonggi provinces within days. At Myeongnyang on the 16th day of the 9th month (26 October 1597), Yi, with thirteen ships, turned back a much larger Japanese fleet and checked its advance to the west. On land the Japanese did not attack Hanseong and soon fell back to their forts on the south coast. The war ended when the last Japanese forces withdrew from Busan in December 1598.

In Japan itself, on 5 February 1597, twenty-six Catholics, Japanese believers and foreign missionaries, were crucified at Nagasaki on the orders of Toyotomi Hideyoshi, the ruler who had launched the invasions of Korea. Their execution marked a hardening of his policy toward Christianity.

At the other end of Eurasia, seafaring powers were also staking everything on fleets. In October 1597 Spain sent a third Armada of about 140 ships against England, but a storm scattered it in the English Channel. In the same year the first Dutch expedition to the East Indies returned home, opening the way for the Dutch spice trade and the later Dutch East India Company.

None of these events was connected with Chilcheollyang. Setting them side by side shows only that, in the same year, the fate of fleets and the control of sea routes mattered in many parts of the world.

Chilcheollyang has long been remembered in Korea as the shadow of Yi Sun-sin’s victories. The Encyclopedia of Korean Culture describes it as the only naval battle the Joseon navy lost in the war, and in popular histories and in television dramas such as KBS’s Immortal Admiral Yi Sun-sin (2004–2005) the defeat sets the stage for Yi’s return, the famous story of the twelve warships he had left, and his victory at Myeongnyang with thirteen.

On Chilcheon Island itself there is now a Chilcheollyang Battle Memorial Park, run by the Geoje Marine Tourism Development Corporation; its exhibition hall shows the strength of the two navies and the course of the battle. Memorial ceremonies for the Joseon sailors who died there are held in Geoje; a regional broadcaster reported one in September 2026.

The commanders are remembered in different ways. Yi Eok-gi, who received the posthumous name Uimin, and An Hong-guk were enshrined beside Yi Sun-sin at Chungminsa in Yeosu, the shrine built in 1601. Won Gyun’s tomb in Pyeongtaek, Gyeonggi Province, was designated a Gyeonggi monument in 1980, and his descendants hold rites there. The records used here do not say that his body was recovered, and Korean Wikipedia mentions a tradition that the tomb holds only his belongings.

The size of the Joseon fleet is uncertain. In the 5th lunar month the Border Defence Council, summarising a report by Gwon Yul, counted 134 large panokseon at Hansan, five or six more on their way and 48 being specially built, to be finished within twenty days: about 180 in all, besides smaller boats. The Encyclopedia of Korean Culture says that Won Gyun left Hansan with “160-odd” ships of the three provinces; Hawley’s history, as cited by English Wikipedia, speaks of up to 200. In the 6th month the Ministry of War told a Ming general that the navy under Won Gyun numbered 4,500 men, a figure that may count only part of the crews. The ship counts are closer to each other than the manpower figures, and the sources used here do not allow a firm total.

For the Japanese, Yi Won-ik reported more than 600 ships arriving at Busan from Japan on the 8th day of the 7th month, and the Encyclopedia’s article on Bae Seol speaks of 600-odd ships at Busan. Hawley, as cited by English Wikipedia, estimates 500–1,000 ships. Neither figure shows how many took part in the attack at Chilcheollyang itself, where the first assault, according to Kim Sik, was made by only five or six ships.

Joseon losses were heavy but are not precisely known. Kim Sik said that all the warships were burned and that the commanders and men were burned or drowned. A Japanese clan record, as quoted in Japanese Wikipedia, claims more than 160 ships captured and thousands of Koreans killed; English Wikipedia, citing Hawley, gives 188 ships destroyed. Yet a report from Gwon Yul, dated the 21st, passed on what his officer Choe Yeong-gil had heard: that 18 large ships had reached Saryang and more than 20 Jeolla ships were scattered along the coast, and that because the fleeing crews had beached their ships, not many rowers had died. The usual figure of “about twelve ships” refers to the squadron Bae Seol kept together; more men and boats may have survived elsewhere.

No securely sourced total for Japanese losses is established here: Japanese Wikipedia lists eight warships sunk without citing a source, and English Wikipedia describes the losses as minimal. The total Joseon loss of life is also unknown; Jingbirok reports the loss of more than 400 officers and men at Gadeok before the battle, without saying how many were killed. In the 8th month Yi Won-ik reported that the accounts of which commanders had died or survived kept changing, and judged that almost all of them, living or dead, had fled rather than fought.

The earliest account is the report of the royal messenger Kim Sik, recorded in the Veritable Records of King Seonjo under the 22nd day of the 7th month. It is short and close to the events, but it gives one survivor’s view; it says that “all” ships were burned and all men died, which later reports qualified. The same day’s record of the audience with the ministers preserves Yu Seong-ryong’s summary of the battle and the king’s reactions. Reports from Gwon Yul, Yi Won-ik and local magistrates in the following weeks add details and contradictions.

Won Gyun’s death is the clearest example of how the sources diverge. Kim Sik saw Japanese soldiers reach Won Gyun but did not see him die. Gwon Yul’s report of the 21st passed on a rumour that Won Gyun had escaped and was heading for Jinju; another report a week later, based on survivors, said that he had landed near Gyeonnaeryang and had surely been killed. On the 5th day of the 8th month the court decided to wait until Won Gyun reappeared before deciding his punishment. Jingbirok says that some said he was killed and others that he escaped, and that the truth was never known; the Revised Annals, compiled in 1657, narrate his death after landing as a settled fact. He never reappeared, and in 1604 he was honoured posthumously.

Most of the vivid details about Won Gyun’s conduct come from later and partisan works. Jingbirok, written after the war by Yu Seong-ryong, Yi Sun-sin’s patron, says that Won Gyun overturned Yi’s arrangements, dismissed Yi’s officers, lived with a favourite concubine in Unjudang, the hall where Yi had held his councils of war, screened off by a double hedge, drank heavily and punished without measure; the Revised Annals also say that he overturned Yi’s regulations, punished without measure and lost the men’s loyalty. The original Veritable Records, compiled under King Gwanghaegun from 1609, contain no such portrait in the entries used here, but preserve Won Gyun’s own reports and the orders sent to him. Even dates differ: Jingbirok gives the defeat under the 7th day of the 8th month, which does not match the court records, and the Encyclopedia of Korean Culture dates the sortie near Busan to the 14th of the 7th month, later than Yi Won-ik’s report.

Japanese sources are fewer in the sources used here. Japanese Wikipedia quotes the Shimazu family documents for the capture of 160-odd ships and the burning of Korean ships along the coast, and the Japanese army general staff’s history for the earlier fight at Gadeok. Such clan records tend to magnify their own achievements, and the sources used here do not allow them to be checked in detail.

For centuries the usual answer was Won Gyun. Jingbirok portrays him as a jealous rival who slandered Yi Sun-sin, took his post and then lost the fleet through drink, harshness and poor judgement, and the Revised Annals also criticise him for overturning Yi’s regulations, punishing without measure and losing the men’s loyalty. This picture shaped later Korean memory, especially as Yi became a national hero from King Jeongjo’s time onward and again in the twentieth century. The Encyclopedia of Korean Culture notes that Won Gyun has often been recorded as the typical cowardly commander.

The contemporary records are more mixed. On the day the defeat was reported, King Seonjo himself said that the fleet should have held Hansan but had been “pressed to go out”, and that the commander-in-chief’s pressure on Won Gyun had caused the defeat. A month later the Office of the Inspector-General went further, saying that the naval defeat “truly came from the failure of the court’s plans”. When Yi Won-ik proposed that Won Gyun bear the penalty for losing the army, the Border Defence Council replied that the fault could not be laid on Won Gyun alone. In 1604 Won Gyun was named a first-class Seonmu merit subject, together with Yi Sun-sin and Gwon Yul.

Since the 1980s some writers, together with Won Gyun’s descendants and local groups in Pyeongtaek, have argued for reappraising him, pointing to his record in 1592, his warnings that the fleet could not safely attack Busan while the Japanese held Angolpo and Gadeok, and the orders that drove him out. Critics answer that such efforts are shaped by family and local loyalty, and that Won Gyun had himself criticised Yi for not attacking. The records discussed here allow his decisions to be weighed alongside the court’s pressure, the exhaustion of the fleet and Japanese control of the coast.

The sources used here point to several causes acting together: a court and a commander-in-chief who demanded an offensive the fleet’s commanders thought impossible; a Japanese army that held the coast and could combine attacks by land and sea; an exhausted and demoralised fleet that anchored in a narrow strait without adequate warning; and decisions by Won Gyun in the last days, including, as Jingbirok reports, his refusal to leave the narrow anchorage despite Bae Seol’s warning. How much weight to give each of these remains a matter of judgement, and the Japanese side of the story is still much less documented in the sources available here.