Operation Ketsugō: Japan’s Preparations for the American Invasion
By August 1945, Japanese commanders had marked three stretches of southern Kyushu as likely American landing grounds. They expected landings on the Miyazaki coast, on the shores of Ariake Bay, and on the western approaches to the Satsuma Peninsula. Their estimate was close to the principal assault areas in Operation Olympic, the first stage of the planned invasion of Japan. Divisions moved toward those coasts. Engineers dug shelters and gun positions while headquarters plotted routes for the reserves. The army and navy assembled aircraft and small attack craft to strike the invasion fleet. By the time Japan surrendered, the defenders had put much of Operation Ketsugō into place along those coasts.1
Kyushu, the southernmost of Japan’s four main islands, was the main target in Operation Downfall’s first phase. Its defenders ranged from experienced formations to new coastal units short of transport and training. Their positions ranged from completed caves to unfinished fieldworks. Ketsugō depended on striking American ships, containing the troops who reached shore, and moving reserves against them before they built a secure base.

A decisive battle on Japanese soil
Japan’s military leaders began serious planning for the defense of the home islands in 1944, as American advances through the central Pacific brought bombers within range and threatened the approaches to Japan. The earlier Sho-Gō plan sought to slow the advance through battles in the Philippines, Formosa, and the Ryukyus. Fighting far from the home islands could inflict losses and buy time for a final defense. The destruction of Japanese air and naval forces, the loss of the Philippines, and the approach to Okinawa made that final battle increasingly likely.2
In January 1945, the army began raising and redeploying dozens of divisions, brigades, and supporting units. On April 8, Imperial General Headquarters issued the outline for Ketsugō, the decisive defense of Japan and nearby territories. Seven regional plans covered possible attacks from Hokkaido to Korea. Ketsugō No. 6 covered Kyushu. There, Second General Army directed Sixteenth Area Army and its three subordinate armies. No. 3 covered the Tokyo region and the Kantō Plain, the likely ultimate target after Kyushu.3
Kyushu received urgent attention because it lay closest to the American bases in the Ryukyus and offered airfields, anchorages, and routes toward the rest of Japan. Japanese officers expected an American landing there in autumn 1945, followed by a larger attack near Tokyo. They also considered attacks on Shikoku, northern Kyushu, or inland airfields. They put forces near all three principal Olympic landing areas while retaining others against attacks elsewhere.4
Japan’s leaders hoped to change American political calculations by inflicting heavy losses on the transport fleet and the troops it put ashore. If the cost of taking Kyushu gave Washington reason to negotiate, they believed Japan could secure terms that avoided occupation. The ground campaign therefore had a political purpose even if American troops succeeded in establishing a beachhead.5
American planners set November 1 for Olympic. They intended to seize southern Kyushu and use it as a base for the larger operation against the Tokyo area. Japanese commanders worked to their own estimates of when the invasion might come. In the spring, Second General Army even allowed for an invasion after late June. That fear encouraged immediate construction and mobilization, while the expectation of an autumn attack gave commanders an October target for finishing the defenses. Japanese divisions were still moving into position and building defenses when the war ended in August.6
The battle before the landing
Japanese planners intended the decisive battle to begin at sea. They concentrated on troop transports, which offered a chance to destroy landing units while their men and equipment were still packed aboard ship. Submarines, conventional aircraft, and special attack units were to find and strike the approach convoys. As the ships closed on Kyushu, suicide aircraft would attack in waves. Midget submarines, manned torpedoes, and small explosive boats were assigned to waters near the anchorages and landing areas. Coastal artillery would fire into sea approaches and against landing craft, while ground artillery shifted toward the beaches once troops came ashore.7
The navy’s Shin’yō boats and the army’s small assault craft drew on Japan’s many coves for concealment. Plans also placed Kaiten manned torpedoes and small submarines near likely routes. Some designs were still being tested, some craft had limited range, and crews had to reach the right waters at the right time. Aircraft posed a similar problem. Crews had to bring them out of hiding, fuel them, and time their sorties so that enough planes reached the transport fleet together to overwhelm its fighter defenses. Japanese planners expected suicide aircraft to destroy a tenth of the transports. That estimate depended on launching the concentrated strike they envisaged.8
The air forces had already learned from Okinawa that suicide attacks could damage a fleet and still fail to stop a landing. For Kyushu they hoped to concentrate more aircraft, including trainers and older types, against shipping packed with troops and supplies. Saving aircraft for that moment reduced training and left the defenders with many inexperienced pilots. Fuel shortages and the dispersal needed to protect airfields made a massed strike harder to organize. Army and navy air units were instructed to cooperate, but their command systems were never fully integrated. The plan required coordinated action from services that had often planned and fought separately.9
The threat of mines and small craft extended beyond the landings themselves, since unloading depended on keeping the bays open. American intelligence was already tracking those preparations. Allied codebreakers followed Japanese troop movements, naval preparations, and radio traffic from new Kyushu commands during the spring and summer. In April, General Douglas MacArthur’s intelligence staff estimated about 230,000 troops on the island and expected further arrivals. Intercepts then identified divisions moving from Manchuria and northern Japan and revealed the formation of additional headquarters in southern Kyushu. By June, the staff had recognized both the Fifty-seventh Army in the south and the Fifty-sixth in the north. Decrypts also revealed that naval depots were converting trainers for suicide missions and that the navy was building explosive boats at Sasebo. The accumulating reports forced American planners to revise their picture of both the garrison and the threats offshore.10

Three armies, several possible battles
Sixteenth Area Army assigned three armies to defend Kyushu. Fifty-seventh Army covered the southeastern approaches, including Miyazaki and Ariake Bay. Fortieth Army covered the southwest and Satsuma Peninsula. Fifty-sixth Army guarded northern Kyushu, its industries, straits, and approaches from the west. Area army headquarters retained forces in central locations from which it hoped to reinforce the decisive sector. Each subordinate army commanded several divisions, roughly the role of a Western corps.11
Within the southern plan, known as Mutsu-Gō No. 1, the Japanese separated three contingencies. A landing near Miyazaki would face coastal divisions positioned on and behind the long eastern shore. An attack into Ariake Bay threatened the Osumi Peninsula and the routes toward Kagoshima Bay. A landing near Kushikino would put Fortieth Army’s western positions under pressure and open another approach toward Kagoshima. Each contingency assigned coastal units and reserve routes for a different main landing.12
There were also plans for the northern and central parts of the island. The northern plan covered the northwest coast between Moji and Sasebo, where Fifty-sixth Army could draw on naval forces around the base. The central plan covered approaches from Shimabara Bay in the west to the Bungo Channel in the east. Headquarters expected attacks on much of that central belt to be small enough for local forces to contain until help arrived. The southern beaches were considered the likeliest place to fight, but the plan gave the north priority if landings struck both ends of Kyushu at once. The southern armies therefore had to plan around demands on the reserve elsewhere on the island.13
The distinction between coastal and mobile divisions was central. Many coastal formations raised in 1945 had little organic transport or heavy artillery. They were expected to fight from prepared positions, obstruct an advance, and hold long enough for stronger formations to assemble. The mobile divisions and independent tank brigades were intended for counterattack. Several veteran formations gave this force a harder core, but newly formed units varied sharply in training and equipment. The division symbols on a map concealed those differences.14
The Miyazaki contingency shows how much the plan asked of those formations. The 154th, 156th, and 212th Divisions were assigned to the eastern coastal defense. If an assault came there, the 25th, 77th, and 216th Divisions, with the 4th, 5th, and 6th Independent Tank Brigades, were to converge as the main counterattack force. The 25th Division and the three tank brigades appeared in the Ariake and Satsuma plans too. Headquarters had to choose their destination. More divisions might arrive from Fifteenth Area Army in a later phase.15
American estimates of Japanese strength on Kyushu rose rapidly through the summer. One estimate put the total at 421,000 on August 1 and was revised to 680,000 shortly before the war ended. A postwar review found about 735,000 personnel across all services, including 25,000 on offshore islands outside the wartime estimate. Part of the rise came from identifying naval recruits and crews from immobilized ships who had been assigned to shore duty. The total included infantry divisions, base personnel, and naval recruits with very different combat roles.16
Mountains and bays separated the landing areas, and few roads ran between them. After the first troops came ashore, Japanese staffs had to decide whether they were seeing the main assault, a supporting attack, or a diversion. The longer they waited, the stronger an American beachhead could become. Moving too early toward the wrong coast would leave the actual main landing with fewer defenders.17
The area army’s August defense map shows a concentration of divisions and tank brigades in the interior, with arrows toward likely southern battlefields. The coastal divisions had to survive the bombardment and keep the invader near the shore while the reserve crossed the interior.18
Artillery had an assignment in both phases of the fight. Guns attached to the coastal divisions were to fire on transports and landing craft. Once a beachhead took shape, the plan called for artillery to leave its coastal positions and support the counterattacking troops. Headquarters envisaged the tanks as fire support for that attack. Commanders also hoped to close with the landing troops so that American ships and aircraft would have difficulty firing without endangering their own men. These orders depended on guns surviving the first bombardment and crews being able to move them as the battle shifted inland.19
Sixteenth Area Army expected its mobile units to cover roughly 15 to 20 kilometers each night and launch a general counterattack within ten days of the landing. Men had to assemble after the landing was identified, move along particular roads, bring artillery and ammunition, and coordinate an attack with formations arriving from other sectors. Night movement reduced exposure to aircraft and limited the hours available for marching. In the meantime, American troops could unload, widen their perimeter, and bring guns ashore. The schedule depended on the defenders keeping the landing force bottled up for several days.20
What a fortified beach meant
Japanese fortification instructions issued in March 1945 described a layered system. Small beach positions would observe and fire on landing craft. Behind them, a foreground zone of concealed weapons, obstacles, mines, and assault routes would slow troops leaving the sand. A main resistance zone of strongpoints, organized for antitank fighting and close combat, would hold key ground. Commanders were also to prepare shelters and assembly positions for the reserves, together with rear holding positions and dummy works to attract American fire. The model sketch in the surviving monograph shows artillery crossfire, antitank points, a main line set back from the shore, and a counterattack zone farther inland.21
The instructions emphasized concealment. Caves and tunnels offered protection from bombs and naval shells. Camouflage hid gun pits and entrances from air reconnaissance. Short fields of fire, reverse-slope positions, and firing points able to strike an advancing force from the flank reduced dependence on large concrete structures at the water’s edge. Antitank obstacles and small teams with explosives were to meet vehicles as they pushed inland. The instructions ordered engineers to begin time-consuming underground works first. They aimed to have the framework in place by July and finish fortifications by October. The work still depended on the labor, material, and terrain available to each local command.22
The tactical emphasis shifted while the works were being built. Earlier defensive experience, especially on Iwo Jima and Okinawa, favored keeping much of the main force away from the beach, where naval guns could crush exposed positions. Japanese commanders then confronted another risk: once American troops secured a lodgment, airfield, and supply area, they had proved difficult to dislodge. On June 6, Imperial General Headquarters issued revised guidance for a decisive counterattack close to the landing site, before the invader could consolidate. Coastal units were still to hold prepared positions while mobile forces made early and repeated attacks against the landing force.23
Fifty-seventh Army’s 156th Division illustrates the arrangement on the Miyazaki coast before the June revision. Infantry squads and antitank weapons near the beach were to obstruct the landing. Most of the division occupied higher ground about two miles inland. Small raiding parties would work between the shore and the main position. Artillery and mortars would fire into the beachhead and its approaches. A second defensive line about three miles behind the main positions screened the divisional reserve. The division planned to commit that reserve if American troops threatened to break out.24
Postwar interviews placed only roughly one-fifth to one-third of the coastal troops directly along the beaches. Others occupied sheltered positions farther back. Japanese officers ranked Ariake’s defenses ahead of Miyazaki’s at surrender and estimated cave and tunnel work at 60 to 90 percent complete by district. Their answers also acknowledged unfinished inland positions.25
Construction under a deadline
In late June and early July, an inspection found Ariake Bay’s fortifications around one-third complete and other sectors further behind. By late July, Sixteenth Area Army reported 80 percent of base construction finished and 70 percent of supplies acquired. A later assessment estimated overall construction at about 90 percent by surrender, with most unfinished projects in Fortieth Army’s western sector. The figures cover different dates, areas, and kinds of work.26
Concrete and steel were scarce. In the peak construction months from April through June, the Kyushu program had about 21,200 metric tons of concrete and 1,000 metric tons of steel available. The army allotted more than three million cubic feet of lumber for tunnels and field fortifications. Where the ground allowed it, men dug into hillsides to make protected positions without pouring large amounts of concrete. Timber supported tunnels and helped reinforce earth walls. These choices could produce substantial protection in favored ground, but the results depended on geology, labor, drainage, weapons, and access to each site. A fieldwork of timber and earth could hide a gun from aerial observation. A deeper shelter offered men a chance to survive sustained bombardment, provided they could still reach their firing positions when it lifted.27
Supply was just as uneven. Southern Kyushu received priority, and Second General Army pushed ammunition and equipment south while there was still time. A May emergency transportation test moved 80 percent of its scheduled cargo, which excluded arms and ammunition. Units received a poorly balanced mix of supplies and struggled to unload and disperse it. By August the Fifty-seventh Army had ammunition for about three and a half divisions in sustained combat. It had more than five infantry divisions and two tank brigades. The plan called for line units to receive their full share of ammunition and for the maneuver and attack forces to receive half. In the southwest, the Fortieth Army had about 40 percent of its required ammunition and 30 percent of its rations. Its mobile units were short of fuel as well. Japanese planners expected American air attacks to break north–south communications even before the landings, so the armies had reason to put these supplies where they expected to use them.28
By mid-July air attacks on bridges had already forced the railway to improvise. Trains ran at night, and shuttles carried traffic around breaks in the network. Second General Army’s preparations called for underground caches dispersed across the forward areas and relays of wagons and trucks to get supplies to southern troops if the rail lines failed. Putting ammunition into local dumps reduced the need to run trains during the invasion. It also meant that a mobile division redirected to another sector would depend on the stores and transport available along its new route.29
Mobilization produced formations faster than weapons and training could be supplied. The 100- and 300-series divisions generally held coastal positions, while the 200-series divisions were organized for mobile counterattack. The 300-series units had especially little artillery. Experienced soldiers served beside recruits with only weeks of instruction. The expansion put more men near the expected beaches and strained the supply network that had to support their counterattacks.30
Civilian mobilization added labor for construction, transport, agriculture, and local services. The People’s Volunteer Corps embodied the army’s intention to involve civilians in a final fight. Some drilled with bamboo spears and mines, though combat preparations remained local and uneven. Civilian labor was already helping build positions and move supplies.31
The battle that remained a plan
Japanese headquarters had made detailed plans for a battle on the same southern coasts selected for Olympic. American intelligence tracked the buildup before surrender. Kyushu’s defenses had grown rapidly, with large differences between sectors and units.32
Postwar accounts drew on Japanese plans, records, and officers’ recollections. Their agreement may reflect overlapping testimony. Wartime Allied intercepts independently confirm some unit movements, while inspections record the progress of particular defenses. Those records establish preparation more firmly than they predict combat performance.33
For Ketsugō to work as its commanders intended, coastal troops had to contain the landing long enough for reserves to arrive. Every extra day gave American engineers and supply units time to strengthen the beachheads.
What does this mean for the player?
In The Eastern Beaches (HP), a 15-turn scenario in Japan ’45, U.S. I Corps lands at Cord and Chrysler Beaches in the Miyazaki sector. The Historical Plan variant puts a light Japanese screen on the coast and most defenders on better ground inland. Part of the 6th Tank Brigade is due from the west on the morning of the second day. As the Japanese player, look for the road exits that your infantry can hold long enough to give those reinforcements a chance to intervene. A strong position that the Americans can bypass buys less time than a smaller one that covers the road they need.34

Travel Mode helps the Japanese reserves move along roads and exposes them to air interdiction. A strike can disrupt a unit or cost it movement points. Pick a place for the lead units to deploy before you send the column forward. If the first tanks reach an exposed American flank, weigh the chance to attack immediately against the time it will take your infantry and guns to catch up.35
American units can take losses as they come ashore. Those losses are how the game represents attacks by kamikaze aircraft and small craft. Use engineers to clear lanes through the mines and move infantry off the sand to make room for later waves. A quick advance can take a junction before the Japanese reserves reach it. Holding the ground around that junction gives the lead units a way back to the beach if a counterattack comes in from the flank.36
For a longer fight on the same coast, Eastern Beaches: Beaches to Breakout gives I Corps 60 turns to land and push inland. The larger map and longer scenario give the Japanese tank brigades time and space to maneuver. See how far the landing force can extend its perimeter before the defenders concentrate enough strength to challenge it.37
Notes
-
Headquarters, USAFFE and Eighth U.S. Army (Rear), Homeland Operations Record, Japanese Monograph No. 17, pp. 126–30, maps 11–12; Headquarters Sixth Army, G-2, “The Japanese Plans for the Defense of Kyushu,” December 31, 1945, p. 1, abridged in D. M. Giangreco, Hell to Pay, app. B.↩︎
-
John Ray Skates, The Invasion of Japan, pp. 100–102; Steven J. Zaloga, Defense of Japan 1945, pp. 29–32.↩︎
-
Homeland Operations Record, pp. 55–66; Edward J. Drea, In the Service of the Emperor, pp. 145–47.↩︎
-
Skates, The Invasion of Japan, pp. 103–4; Homeland Operations Record, pp. 101–7, 126–30.↩︎
-
Skates, The Invasion of Japan, pp. 101–4; Giangreco, Hell to Pay, ch. 7. These works draw partly on the same Japanese planning record.↩︎
-
Drea, In the Service of the Emperor, pp. 151–55; Homeland Operations Record, app. III, pp. 157–58.↩︎
-
Homeland Operations Record, pp. 62–67, 127–28; Zaloga, Defense of Japan 1945, pp. 33–35; Sixth Army G-2, “Japanese Plans,” p. 1 and questions 19–23, in Giangreco, app. B.↩︎
-
Zaloga, Defense of Japan 1945, pp. 37–46; Skates, The Invasion of Japan, pp. 108–13; Sixth Army G-2, “Japanese Plans,” p. 1, in Giangreco, app. B.↩︎
-
Skates, The Invasion of Japan, pp. 107–10; Drea, In the Service of the Emperor, pp. 150–51; Homeland Operations Record, pp. 65–67.↩︎
-
Drea, In the Service of the Emperor, pp. 155–57. Allied signals intelligence provides a distinct check on movement and command activation, though it could not measure the combat readiness of each unit.↩︎
-
Homeland Operations Record, pp. 99–102, 122–28, maps 11–12; Giangreco, Hell to Pay, ch. 7, explanation of Japanese formation terms.↩︎
-
Homeland Operations Record, pp. 126–31; Skates, The Invasion of Japan, pp. 119–22.↩︎
-
Homeland Operations Record, pp. 126, 129–32.↩︎
-
Drea, In the Service of the Emperor, pp. 146–49; Zaloga, Defense of Japan 1945, pp. 32–34.↩︎
-
Homeland Operations Record, pp. 126–30.↩︎
-
Headquarters Sixth Army, G-2, “Japanese Plans,” p. 1, in Giangreco, Hell to Pay, app. B; Drea, In the Service of the Emperor, pp. 161–65; Zaloga, Defense of Japan 1945, p. 50. The figures differ in date, scope, and inclusion of naval, service, and offshore personnel.↩︎
-
Homeland Operations Record, pp. 126–31; Skates, The Invasion of Japan, pp. 119–24.↩︎
-
Homeland Operations Record, pp. 127–30, map 12; Drea, In the Service of the Emperor, pp. 151–53.↩︎
-
Homeland Operations Record, p. 127; Drea, In the Service of the Emperor, p. 152.↩︎
-
Drea, In the Service of the Emperor, p. 152; Skates, The Invasion of Japan, pp. 119–24.↩︎
-
Homeland Operations Record, app. III, pp. 157–67; Zaloga, Defense of Japan 1945, pp. 35–37.↩︎
-
Homeland Operations Record, app. III, pp. 157–65; Zaloga, Defense of Japan 1945, pp. 35–37.↩︎
-
Drea, In the Service of the Emperor, pp. 149–53; Skates, The Invasion of Japan, pp. 112–15; Zaloga, Defense of Japan 1945, pp. 35–37.↩︎
-
Drea, In the Service of the Emperor, pp. 151–52.↩︎
-
Sixth Army G-2, “Japanese Plans,” question 31, in Giangreco, Hell to Pay, app. B.↩︎
-
Zaloga, Defense of Japan 1945, pp. 47–48, describes the late June–early July inspection. Drea, In the Service of the Emperor, pp. 148–49, gives Sixteenth Area Army’s late July figures. Giangreco, Hell to Pay, ch. 14, estimates about 90 percent complete by surrender. Sixth Army G-2, “Japanese Plans,” question 31, in Giangreco, app. B, gives district figures for cave and tunnel work. These measures have different dates and denominators.↩︎
-
Zaloga, Defense of Japan 1945, pp. 47–49; Homeland Operations Record, app. III, pp. 157–64.↩︎
-
Homeland Operations Record, pp. 132–36; Drea, In the Service of the Emperor, pp. 147–49; Sixth Army G-2, “Japanese Plans,” question 32, in Giangreco, app. B.↩︎
-
Drea, In the Service of the Emperor, pp. 147–48.↩︎
-
Drea, In the Service of the Emperor, pp. 145–49, 164–65; Skates, The Invasion of Japan, pp. 105–6.↩︎
-
Drea, In the Service of the Emperor, pp. 147–48; Skates, The Invasion of Japan, pp. 129–31; Zaloga, Defense of Japan 1945, pp. 36–37.↩︎
-
Homeland Operations Record, pp. 126–36; Drea, In the Service of the Emperor, pp. 159–65; Skates, The Invasion of Japan, pp. 129–31.↩︎
-
Homeland Operations Record, introduction and pp. 122–30; Sixth Army G-2, “Japanese Plans,” in Giangreco, Hell to Pay, app. B, editorial note. The reproduction contains eighteen of the original forty-eight pages. See also Drea, In the Service of the Emperor, pp. 155–65, and Zaloga, Defense of Japan 1945, pp. 47–50.↩︎
-
Wargame Design Studio, Japan ’45, “The Eastern Beaches (HP),” scenario briefing and reinforcement schedule (#45_1104_6HP); Wargame Design Studio, Design and Campaign Notes, p. 33.↩︎
-
Wargame Design Studio, Panzer Campaigns User Manual, pp. 29–31, 100.↩︎
-
Wargame Design Studio, Design and Campaign Notes, pp. 7, 11–12; Wargame Design Studio, Panzer Campaigns User Manual, pp. 87, 153.↩︎
-
Wargame Design Studio, Japan ’45, “Eastern Beaches: Beaches to Breakout,” scenario briefing (#45_1104_7); Wargame Design Studio, Design and Campaign Notes, p. 34.↩︎
Bibliography
Drea, Edward J. In the Service of the Emperor: Essays on the Imperial Japanese Army. University of Nebraska Press, 1998.
Giangreco, D. M. Hell to Pay: Operation Downfall and the Invasion of Japan, 1945–1947. Updated and expanded ed. Naval Institute Press, 2017.
Headquarters Sixth Army, Office of the Assistant Chief of Staff, G-2. “The Japanese Plans for the Defense of Kyushu.” December 31, 1945. Abridged in Giangreco, Hell to Pay, appendix B. Digital catalog record.
Headquarters, USAFFE and Eighth U.S. Army (Rear). Homeland Operations Record. Japanese Monograph No. 17, incorporating Nos. 17–20. Distributed by the Office of the Chief of Military History, Department of the Army, n.d. Digital scan.
Skates, John Ray. The Invasion of Japan: Alternative to the Bomb. University of South Carolina Press, 1994. First paperback ed., 2000.
Wargame Design Studio. Panzer Campaigns: Japan ’45—Operation Olympic. Version 4.05.3h2. Windows game. Game and scenario documentation.
Wargame Design Studio. Panzer Campaigns: Japan ’45—Operation Olympic. Design and Campaign Notes. Design notes.
Wargame Design Studio. Panzer Campaigns User Manual. Series manual.
Zaloga, Steven J. Defense of Japan 1945. Illustrated by Steve Noon. Osprey Publishing, 2010.




Leave a comment