The Soviet Entry into the War against Japan

The Road to Yalta: Forty Years of Hostility and One Treaty

When Stalin, Roosevelt, and Churchill assembled at the Livadia Palace in Crimea on 4 February 1945, forty years of history were piled upon the question of Soviet entry into the war against Japan.

In 1905, Imperial Russia lost a war to Japan. The Treaty of Portsmouth stripped away southern Sakhalin and the Port Arthur lease. The humiliation outlasted the Tsar. Then, in 1931, Japan seized Manchuria and the following year erected the puppet state of Manchukuo. The Soviet Union and the Japanese Empire now faced each other across 5,000 kilometres of direct border. In 1935, Stalin sold the Chinese Eastern Railway to Japan simply to reduce friction.

Two clashes followed in 1938 and 1939. The Battle of Lake Khasan in 1938 was limited. Khalkhin Gol, from May to September 1939, was not. Georgy Zhukov's 1st Army Group deployed 498 tanks and 557 aircraft to envelop and annihilate the Kwantung Army's 23rd Division. Japanese dead: 8,629. The defeat shattered the Imperial Japanese Army's 'Strike North' faction. Strike South prevailed; Japan would avoid another war with the USSR at almost any cost.

On 13 April 1941, Japanese Foreign Minister Yosuke Matsuoka and Soviet Foreign Commissar Vyacheslav Molotov signed the Soviet-Japanese Neutrality Pact in Moscow. Stalin made an extraordinary personal appearance at the railway station to see Matsuoka off, one of the rare occasions he appeared before a foreign diplomat. When Germany invaded the USSR two months later, Japan honoured the pact and attacked Pearl Harbor instead. Stalin transferred Siberian divisions to defend Moscow. For four years the two powers maintained an eerie coexistence: each fighting the other's allies, neither fighting the other.

At the Tehran Conference in November 1943, Stalin gave his first verbal commitment to enter the war against Japan after Germany's defeat. By October 1944, when an Anglo-American delegation visited Moscow, he was discussing concrete terms. By February 1945, everything had shifted. The Kwantung Army still counted 700,000 men on paper, but its elite divisions had been drained away to the Pacific and China. The remaining force was filled with reservists, draftees, and militia. Most tanks were 1930s models; few soldiers had adequate training. It had never properly recovered from Zhukov's blow in 1939.

The United States was desperate. Okinawa had not yet been fought, but the arithmetic of Operation Downfall, the planned invasion of the Japanese home islands, already projected American casualties ranging from 250,000 to a million. The Joint Chiefs assessed that unless Soviet forces pinned the Kwantung Army in Manchuria, Japan could concentrate its continental forces on homeland defence. For Roosevelt, Soviet entry was not a bonus; it was a strategic necessity that might save hundreds of thousands of American lives.

All of this lay on the table at Yalta. Stalin knew what Roosevelt needed, and he knew what he could demand in return. The statute of limitations on 1905 had not yet expired.

A Promise in February, a Signal in April: The Yalta Secret Pact and the Denunciation

At Yalta (4-11 February 1945), the official agenda was the postwar settlement of Germany. But Roosevelt's real urgency was Japan. The atomic bomb had not yet been tested; the planned invasion of the home islands, Operation Downfall, projected hundreds of thousands of American casualties. The US Joint Chiefs of Staff estimated that without Soviet entry, victory could stretch beyond 1946. Roosevelt arrived in Crimea already barely able to move, and made securing Stalin's commitment to the Pacific War his overriding priority.

Stalin knew exactly what leverage he held. In December 1944, he had spread a map before US Ambassador Averell Harriman in Moscow and laid out his desiderata: southern Sakhalin and the Kuril Islands, restoration of the Russian leasehold at Port Arthur and Dairen, joint operation of the Manchurian railways, and preservation of the status quo in Outer Mongolia. At Yalta, he got every item on that list. The "Agreement Regarding Japan," signed on 11 February, contained three provisions: Outer Mongolia's status quo; restoration of Russian rights "violated by the treacherous attack of Japan in 1904" (southern Sakhalin, Port Arthur, Dairen, the Manchurian railways); and transfer of the Kuril Islands. But every word was strategic. The Kurils, never Russian before 1904, could not be claimed as a "restoration," so Stalin placed them in a separate article using "handed over" rather than "returned." And by securing Roosevelt and Churchill's pledge that these claims "shall be unquestionably fulfilled after Japan has been defeated," he preemptively blocked any later challenge under the Atlantic Charter or the Cairo Declaration.

The agreement ran roughshod over Chinese sovereignty, but Chiang Kai-shek was not a party to it. Roosevelt merely promised to "take measures" to obtain Chiang's concurrence "on advice from Marshal Stalin." Korea's fate was discussed on 8 February in a private Roosevelt-Stalin conversation without Churchill present: a 20-to-30-year trusteeship was floated. But it produced no line in the final protocol. The secret agreement itself remained hidden from the public until 1947.

On 5 April 1945, at 3 p.m., the Soviet Union sent its first open signal. Molotov summoned Japanese Ambassador Satō Naotake to the Foreign Commissariat and informed him that the USSR would not renew the 1941 Neutrality Pact. The statement declared that "the situation has been basically altered": Japan, Germany's ally, was aiding Germany's war against the USSR, and was fighting the USSR's own allies, the United States and Britain. "In these circumstances the neutrality pact between Japan and the USSR has lost its sense, and the prolongation of that pact has become impossible."

Satō immediately asked: was the pact being annulled forthwith, or would it remain in force until its expiry in April 1946? Molotov confirmed the latter. Under Article Three, a denunciation filed one year before expiry kept the pact valid through the end of its five-year term. Stalin had no wish to be seen replicating Hitler's violation of the Molotov-Ribbentrop Pact.

But the real meaning was unmistakable. Time magazine, under the headline "So Sorry, Mr. Sato," noted that Molotov's tone suggested "this technicality might be brushed aside at Russia's convenience." The Soviet command had already ordered Far Eastern forces onto heightened alert, and the massive transfer of troops and matériel from Europe was underway. The pact was legally alive for one more year but politically dead. What was promised at Yalta in February and telegraphed in April was now moving inexorably toward August.

One Side Waiting, the Other Racing: The Three-Month Asymmetric Contest

When Germany surrendered on May 9, 1945, the Yalta secret protocol's clock started ticking. Stalin accepted a deadline of August 9 to enter the war against Japan, but those three months were anything but a waiting period. The USSR had to redeploy vast forces from Europe to the Far East, while Japan, paradoxically, was racing toward that same Soviet Union to ask it to mediate an end to the war.

The troop movement was the largest strategic redeployment of World War II. From May through early August, over 400,000 troops, 7,137 guns and mortars, and 2,119 tanks and self-propelled guns were shipped east on the Trans-Siberian Railway. Twenty to thirty trains per day ran toward the Far East. The most dramatic movement was that of the 6th Guards Tank Army, stationed in Czechoslovakia. It traveled 9,000 km to Choibalsan in Mongolia aboard 88 trains of 60 cars each; the first elements arrived on July 17, and the entire transfer took 30 days. At Choibalsan, 185 American Lend-Lease M4A2 Sherman tanks awaited them. In parallel, at Cold Bay, Alaska, Project Hula secretly transferred 149 US Navy vessels to the Soviet Navy and trained 12,000 Soviet naval personnel.

Japan was moving toward the Soviet Union in an entirely different direction. As defeat at Okinawa loomed, Emperor Hirohito and the Supreme War Council pinned their last hope on Soviet mediation to end the war. On June 7, 1945, former prime minister and foreign minister Hirota Kōki made unofficial contact with Soviet ambassador Yakov Malik. Hirota sounded out the possibility of renewing the Neutrality Pact or even concluding an alliance. But Molotov instructed Malik to "neither agree to specific conditions nor cut off negotiations." Malik deliberately slowed the talks by sending Japanese proposals through the diplomatic pouch. On July 12, Foreign Minister Tōgō Shigenori instructed Ambassador Satō Naotake to deliver an imperial message requesting mediation, but Molotov, preparing to depart for Potsdam, refused to meet Satō.

In Moscow, meanwhile, a decisive shift occurred. When Harry Hopkins visited as Truman's special envoy in late May, Stalin promised that Soviet forces would be in "sufficient preparedness" by August 8, though he qualified that the actual start date depended on weather. At this meeting, Stalin for the first time raised the question of occupation zones in Japan. Hopkins promised to place a joint ultimatum to Japan on the Potsdam agenda, giving Stalin precisely the justification he needed to enter the war in violation of the Neutrality Pact.

On June 26 and 27, Stalin convened a combined meeting of the Politburo, government, and military leadership, which made the final decision to launch an all-out offensive in August. The General Staff's plan for three fronts attacking simultaneously toward the center of Manchuria was approved. The question of occupying Hokkaido was fiercely debated and left unresolved. Meretskov and Khrushchev argued for occupation; Voznesensky, Molotov, and Zhukov opposed it as too risky and likely to provoke US counteraction. Stalin asked Zhukov how many additional divisions would be needed. "At least four," Zhukov replied. Stalin said nothing further. On June 28, Stalin issued three directives to the Far Eastern front commanders: complete all attack preparations by July 25 or August 1.

At the Potsdam Conference (July 17 to August 2), the framework of the war shifted again. On July 26, the United States, Britain, and China issued the Potsdam Declaration demanding Japan's unconditional surrender. The Soviet Union was not among the signatories: it was still neutral toward Japan. Stalin was furious at this exclusion. According to Tsuyoshi Hasegawa's research, Stalin concluded that the United States intended to end the war without the Soviet Union. When Truman casually informed Stalin on July 24 that the US possessed a "powerful new weapon," Stalin reacted calmly; Soviet intelligence had already penetrated the Manhattan Project. During the conference, Stalin telephoned Vasilevsky and asked whether the attack could be moved up by ten days from the August 11 date set by the Stavka. Vasilevsky answered that the concentration of troops and supplies made this impossible. Stalin, for the time being, backed down. The fear that the war might end before the Soviet Union could join it dominated every calculation Stalin made.

Forty-eight Hours Pulled Forward: From Hiroshima to the Declaration of War

The attack date had originally been set for 11 August. On 3 August Vasilevsky reported to Stalin that the Transbaikal Front had advanced to within fifty to sixty kilometres of the border and would be ready by the morning of 5 August, but that the 'optimal time' for crossing the border was 9-10 August. Because surprise required all fronts to strike on the same day at the same time, the Stavka rejected the proposal and left the attack time at midnight Chita time on 11 August. But then the situation turned upside down at Potsdam.

On 24 July, during a recess, Truman approached Stalin and obliquely mentioned that the United States had acquired 'a new weapon of unusual destructive force.' Stalin appeared indifferent, but thanks to his spies inside the Manhattan Project he already knew. A greater shock came on 26 July. Resolved to exclude the USSR, Truman and Byrnes had deleted from Stimson's draft both the clause allowing Japan to keep the monarchy and all references to Soviet entry, then released the Potsdam Declaration to the press before sending Molotov a copy as a 'diplomatic courtesy.' When Molotov asked for the release to be delayed two or three days, Byrnes said it was too late. When Stalin asked to sign, Washington fobbed him off with Paragraph 5 of the 1943 Moscow Declaration and Articles 103 and 106 of the UN Charter, documents that had nothing to do with Japan. Byrnes later recalled: 'I would have been satisfied had the Russians determined not to enter the war.'

Truman sought to force Japan's capitulation with the atomic bomb before the Soviets came in; Stalin sought to enter before Japan surrendered. It was a race to the finish. Then, on 6 August, the bomb fell on Hiroshima. The news seems to have crushed Stalin. Pravda carried nothing about Hiroshima on 7 August and only on 8 August printed Truman's statement, without comment, at the bottom of page four. Stalin saw no one in his Kremlin office on 6 August. He seems to have believed the game was over, that Japan would surrender and the Americans had won. When Molotov met the Chinese delegation led by T. V. Soong on the afternoon of 7 August, he was dejected and told Soong that Japan was on the verge of collapse.

But Hiroshima did not bring Japan's surrender decision. On 7 August Foreign Minister Tōgō Shigenori sent Satō another urgent cable, ordering him to seek an immediate meeting with Molotov to get Moscow's answer about the Konoe mission. Japan, despite the bomb, still clung to Soviet mediation. When Satō's request arrived that afternoon, Stalin leapt into action. At 4:30 p.m. he ordered Vasilevsky to launch the Manchurian operation the next day, at midnight Chita time on 9 August (6 p.m. Moscow time on 8 August), moving the attack up by forty-eight hours.

Vasilevsky issued four directives between 10:35 and 11:10 p.m., ordering all three fronts and the Pacific Fleet to begin operations at 6 p.m. Moscow time on 8 August. That same night, when his differences with the Chinese delegation narrowed no further, Stalin decided to enter the war without any agreement with China. At 5 p.m. on 8 August Molotov cut short Satō's greeting and read the declaration of war. Within an hour, Soviet tanks and aircraft had crossed the border. Stalin had, in the end, joined the war in the nick of time.

Once the Border Was Crossed, the Kwantung Army Was Already Fighting the Wrong War

The shock of 9 August lay less in the numbers than in the form of the war. Japan expected the Soviets to batter their way through the seventeen fortified border areas one after another, then advance slowly along roads and railways. The Kwantung Army's spring plan was to delay the offensive with forward units, hold the Soviet forces in central Manchuria, and counterattack with reinforcements from China and Korea, holding out for at least a year. But that plan was already hollow. Since 1945, elite divisions and equipment had been transferred to the Pacific and China fronts. The units left behind contained many raw recruits, obsolete tanks, and inadequate antitank guns. The puppet armies of Manchukuo and Mengjiang added numbers but could not repair the gaps in command and combat power.

The Soviets did not attack this weakness frontally alone. Vasilevsky's design was a deep envelopment: Malinovsky's Transbaikal Front and the 1st Far Eastern Front under Meretskov would converge on central Manchuria, while the 2nd Far Eastern Front crossed the Amur from the north and pressed the rear. The 6th Guards Tank Army and Issa Pliyev's Soviet-Mongolian cavalry-mechanized group chose routes across the Gobi and Greater Khingan rather than the obvious roads. The choice turned terrain the Japanese considered impassable into an operational opportunity. Tanks and vehicles depended on fuel and water columns following behind, but success would cut the Kwantung Army's border defenses from its central command in one movement.

In the east, Meretskov's forces widened their breaches instead of becoming fixed in the fortified zones. In the north, the Amur crossings, supported by the river flotilla, restricted Japanese movement. The assault of 9 August was therefore not merely an attack from three directions. Soviet aircraft, radio, armored spearheads, and forward command posts were coordinated around objectives hundreds of kilometers beyond the frontier, toward Changchun and Harbin. Soviet operational art, learned in the war against Germany, was applied to Manchuria: not simply taking a trench line, but successively paralyzing the enemy's command, communications, and reserves.

Mongolia's entry meant more than a legal formality. The Mongolian People's Republic declared war on 10 August, but its units had already crossed the frontier the day before as part of the Soviet-Mongolian group. From Ulaanbaatar's perspective, this was a struggle to reverse the consequences of the border conflicts of the 1930s, confront the Japanese puppet state of Mengjiang and Manchukuo, and secure postwar safety alongside the Soviet Union. Mongolian cavalry and motorized units provided reconnaissance and flank protection across steppe and desert. The price was that the postwar order would be reorganized under overwhelming Soviet military and political influence. The border war Japan hoped would buy time thus became, from its first hours, a war in which the rear and flanks were disappearing. The Kwantung Army was not facing one army crossing a line, but three operational spaces closing at once.

The treaty promised Manchuria, but surrender split the military

Signed in Moscow on 14 August 1945 by Wang Shijie and Vyacheslav Molotov, the Sino-Soviet Treaty of Friendship and Alliance was less a wartime alliance than a prearrangement for Manchuria's postwar order. The price accepted by the Nationalist government was substantial. If a plebiscite were held, China would recognize Outer Mongolia's independence; the Chinese Eastern and South Manchurian railways, combined as the Chinese Changchun Railway, would be jointly owned and operated by China and the Soviet Union for thirty years; Port Arthur would become a joint naval base; and Dalian would be a free port in which the Soviets held preferential rights. In return, Moscow promised to withdraw from Manchuria within three months after expelling Japan and to respect Chinese sovereignty. The Soviet account recorded by the US State Department on 17 August also said that a representative of the National Government would administer the territory after it was cleared of the enemy. The Nationalist calculation was plain. To end Japan's military rule and obtain international recognition of Chinese territorial sovereignty, it had to bargain with Soviet forces already entering Manchuria. Yet the document guaranteed sovereignty in words and eventual withdrawal, not who would control weapons, railways, and cities on the ground.

On the same day, Japan's Supreme Council for the Direction of the War was conducting a different kind of bargain. Even after Hiroshima, Army Minister Korechika Anami and Army Chief of Staff Yoshijiro Umezu maintained that Japan should fight a decisive battle at home, inflict enormous losses on the United States, and then secure better terms. Foreign Minister Shigenori Togo and Navy Minister Mitsumasa Yonai leaned toward accepting the Potsdam terms, but insisted that the imperial institution must be preserved. Soviet entry destroyed the premise of both the diplomatic and military calculations. The Soviet Union, which Japan had tried to retain as a mediator, was now attacking in Manchuria, and the northern front eliminated the Soviet neutrality assumed by the plan for a decisive battle in the Home Islands. Even so, the council did not immediately agree. Japan had already conveyed its willingness to accept on 10 August, but the condition concerning the Emperor remained unresolved, and on the night of 14 August some military hardliners tried to prevent the surrender broadcast. The cabinet of Kantarō Suzuki finally followed the Emperor's decision, and Hirohito announced acceptance of the Potsdam Declaration in a recorded address. At noon on 15 August, most people heard the Emperor's voice for the first time. He avoided the plain words “defeat” and “surrender,” saying instead that he would “bring the war situation to a conclusion.” The phrasing protected military dignity and institutional continuity, but obedience did not automatically follow. That is why some Kwantung Army units continued fighting. Political surrender in Tokyo and military control in Manchuria did not occur at the same moment. In that interval, the reality of occupation began to shape borders and power before the treaty's promises could do so.

The Surrender Order Was Not the End: From Manchuria to Missouri

The surrender order of 16 August was not a magic document that stopped the entire Kwantung Army at once. The Emperor's broadcast was not a direct order delivered to every commander at the front, while Soviet disruption of communications and the speed of the advance made it difficult for headquarters at Hsinking to retransmit instructions. The army itself was divided. General Yamada Otozō and part of the command accepted the Emperor's decision as the basis of military obedience and chose to cease fire. Some staff officers and field units, however, still believed that a final fight might secure better terms. To them, surrender meant not simply defeat but the collapse of colonial rule, the army's honor, and the danger that their soldiers would become prisoners. In the conditions created by broken communications, that argument produced not an operation but the isolation of individual units. After 17 August the surrender order was transmitted repeatedly, and Soviet forces seized such centres as Harbin and Hsinking. Fighting continued in some places until the end of August, but there was no longer a Japanese force capable of directing Manchuria as a whole. The cost did not disappear with the order. Hundreds of thousands of soldiers and civilians were left amid disrupted movement and shortages, while many Kwantung Army soldiers became Soviet prisoners and were mobilized for prolonged labor. Thus the accounts that treat 15 August as the practical end of the war and those that stress continuing fighting and coercion in Manchuria and Northeast Asia are not mutually exclusive. The first describes the decision in Tokyo; the second describes the time it took that decision to reach the ground. The Instrument of Surrender, signed aboard USS Missouri on 2 September, legally closed this gap. Japan's government and Imperial General Headquarters accepted the Potsdam Declaration, proclaimed the unconditional surrender of “the Japanese Imperial General Headquarters and of all Japanese armed forces and all armed forces under Japanese control wherever situated,” and commanded the Imperial Headquarters to issue immediate surrender orders to every formation. The document was signed by command of the Emperor, yet it also placed the authority of the Emperor and Japanese Government under the Supreme Commander for the Allied Powers. The signature of the Soviet representative, Kuzma Derevyanko, made clear that the surrender in Manchuria was not an American victory alone but a capitulation to the Allied Powers. Yet ending combat and creating an equitable occupation order were different matters. Surrender ended Japan's military sovereignty, while the future of Manchuria and Korea remained contested among occupation forces and Chinese political movements.

The Empire Fell, the Frontiers Remained: What the Operation Settled and What It Left Open

The operation settled a military fact above all: Japan could no longer preserve Manchukuo, its puppet state, or the Kwantung Army as a shield for its continental empire. The Instrument of Surrender of 2 September 1945 required the disarmament of Japanese forces and established Allied occupation; the Soviet Union took control of southern Sakhalin and the Kuril Islands. Yet the instrument did not transfer Manchuria's future government or Korea's political sovereignty to Moscow. The 14 August Sino-Soviet Treaty of Friendship and Alliance regularized Soviet rights through an agreement with the Nationalist government, but did not guarantee that the Chinese Communist Party would later exercise state power there. Who filled the vacuum depended not on the surrender document, but on the timetable of occupation and the organizational strength of armed political forces.

That distinction was clearest in Manchuria. Until its withdrawal in 1946, the Soviet army controlled railways and industrial installations, while Japanese weapons and some instruments of local administration passed to the Chinese Communists. This did not create the Communist victory by itself. The Chinese Communist Party already possessed rural base areas and organized armies, while the Nationalist government had been exhausted financially and administratively by the war. Soviet occupation nevertheless changed the geographic conditions and the supply of weapons with which the contest began. In Korea, too, Soviet occupation north of the 38th parallel and American occupation to the south turned liberation from the Japanese empire into a competition between two occupation regimes to build states. The operation ended an empire, but it did not return sovereignty in the liberated spaces to a single political actor.

The most persistent dispute concerns what decided Japan's surrender. Tsuyoshi Hasegawa argues from diplomatic records and meeting materials that the declaration of war and invasion arrived precisely as Japanese leaders were hoping to use the Soviet Union as a mediator. In this interpretation, the atomic bomb at Hiroshima was shocking, but Soviet entry destroyed both Japan's last diplomatic channel for seeking a conditional settlement with the United States and the continental resources represented by Manchuria. Other scholars see the destruction of Hiroshima and Nagasaki, the naval blockade, and accumulated military defeat as pressures that worked together. As Richard B. Frank emphasizes, the Supreme Council for the Direction of the War had still not reached immediate agreement on 9 August; the Emperor's intervention, military resistance, and negotiations over preserving the imperial institution delayed surrender.

The cautious conclusion, therefore, is not to identify one final blow with certainty. Soviet entry annihilated the Kwantung Army and destroyed the mediation strategy on which Japan had relied; the atomic bombings and blockade altered the cost and speed of a collapse already underway. Much of Japan's wartime record was destroyed or concealed, while Soviet materials remained restricted for decades, so no single document proves a decisive causal chain. What is certain is that the operation ended a war while opening the frontiers of the postwar order. The status of the Kuril Islands, the division of Korea, and the reordering of power in Manchuria were not problems solved by surrender. They were new struggles created by it.