Home / Uncategorized / March 18, 1945: The Okinawa Campaign Begins

March 18, 1945: The Okinawa Campaign Begins


A Japanese Judy burning after being shot down by anti-aircraft fire from the USS Wasp (CV-18) off the Ryukyus on 18 March 1945.
A Japanese Judy burning after being shot down by anti-aircraft fire from the USS Wasp (CV-18) off the Ryukyus on 18 March 1945.

This article was published in the May 1954 issue of Proceedings magazine as “Kamikazes and the Okinawa Campaign” by Rear Admiral Toshiyuki Yokoi, former Imperial Japanese Navy, with the assistance of Roger Pineau.

Japan’s special air attack units (Kamikaze) were initially organized under very particular circumstances and with limited operational objectives in the Philippines late in 1944. In the first stage Admiral Ohnishi certainly did not conceive of either allocating more than 24 planes for such suicide attacks or continuing this type of operation indefinitely, because there are serious basic defects in this type of attack. First, the expenditure of life and materiel is great. It takes several years to train one good pilot, yet in Kamikaze operations, he, as well as his plane, will be expended in a single sortie. This runs counter to the most important problem of an operation staff, which is to attain objectives with the least possible expenditure of life and materiel. Second, the striking velocity of a plane is not great enough to penetrate the decks of fleet carriers or battleships and cause critical damage below. A suicide attack on a carrier deck will not strike a vital blow unless the deck is full of planes. Third, operational command of Kamikaze planes is difficult because results cannot be evaluated with any accuracy. When his subordinates’ lives are sacrificed, a commander will naturally tend to overestimate the results achieved. When such overestimates are compounded, a totally erroneous picture will be presented to the high com­mand, whose judgment and decisions in turn will be falsely influenced.

These factors provide substantial reason why wise commanders were opposed to suicide air attacks, and yet the early reports of the Kamikazes’ amazing success caught the fancy of military leaders as well as the public—and the craze was on. The fact that sunken U. S. escort carriers were reported as standard fleet carriers was completely unknown and unrecognized in the surge of enthusiasm which overrode all defects of the Kamikaze attacks.

             Another factor contributing to the situation was the vanity of heroism.

             Thus came the age of suicide air attacks.

Imperial General Headquarters was so fully convinced that it issued an outrageous and unprecedented order to the effect that all armed forces should resort to suicide attack. This proved that the High Command, utterly confused by a succession of defeats, had lost all wisdom of cool judgment and had degenerated to the point of indulging in wild gambling. The order was nothing less than a national death sentence. Like every military order, it was issued in the name of the Emperor and was, therefore, no matter how outrageous, not open to question or criticism. Obedience was imperative; there was no alternative. Critics of the Kamikaze attacks should distinguish the completely volunteer flights of October, 1944, from those made after this Imperial order.


Once the Philippines were lost the Japanese Navy expected the next enemy thrust to be made at Okinawa. Here there were anchorages adequate for a large fleet and sites to accommodate a group of large airfields.

Okinawa is 1,000 miles from Leyte and 1,200 miles from the Marianas, too great a range for land planes to operate, so the United States Navy would probably have to commit all of its carrier striking force in order to cover the landings. On the other hand, it should be comparatively easy for the defending Japanese to maintain supply lines to the Japanese mainland, only 350 miles distant, especially since there were operational airfields on the intervening islands of Kikai and Minami Daito. These circumstances promised a good opportunity to bring air power to bear in striking a serious blow at the enemy task forces.

In early November of 1944 I was given command of the 25th Air Flotilla, which was based at Kanoya in southern Kyushu. A study of recent Japanese defeats had convinced me that the ineffectiveness of our land-based air strength was attributable, in the main, to poor defense measures and the inexperience of the pilots. I felt firmly that carrier-based aircraft were no match for land-based planes properly employed. In the Marianas, the Philippines, and Formosa, our planes had been defeated easily because of inadequate patrol systems and improper defensive tactics generally. Successful operation of land bases demanded improved patrol methods, elimination of surprise by the enemy, and effective measures to protect ground installations so that fighting strength would not be interrupted. If these could be achieved, our shorter lines of communication would provide a supply advantage which would enable us to await an enemy blunder that might lay him open to a fatal blow.

In anticipation of the enemy attack on Okinawa, I concentrated on building up airfield defenses in southern Kyushu. It was fortunate that most of these bases were located on tablelands where tunnels could easily be dug in the sandy earth. Materials and manpower were insufficient for the preparation of underground hangars, but most of the other important installations, such as the command communications system, telegraph stations, ammunition dumps, repair shops, and living quarters, were dug in. The planes were to be concealed in the surrounding forests and hills. Terrain permitting, runways were to be camouflaged to resemble ordinary roads. Emptied barracks and hangars were left above ground as a decoy to absorb enemy attacks. In carrying out these plans I encouraged and supervised the construction crews and also urged off-duty aviation personnel to assist them, so that by early February of 1945 great progress had been made in the earthworks and I felt confidence in our position.

On February 9 I received orders to report to the Naval General Staff as soon as possible. I left Kanoya immediately in a bomber plane which took me to Atsugi, but it was dark when I reached the residence of the Chief of Naval General Staff. There Vice­Admiral Seiichi Ito, the Vice-Chief, greeted me with the information that I was to be appointed Chief of Staff to the Fifth Air Fleet Commander, Vice-Admiral Matome Ugaki, the man who had been Chief of Staff of the Combined Fleet under the famous Admiral Isoroku Yamamoto at the time of Pearl Harbor.

“What is the mission of this Command?” I asked.

“The Fifth Air Fleet will be composed of approximately 600 planes comprising the eight elite aviation units remaining in the Japanese Navy. Its area of operation will extend from Okinawa eastward to a north­south line through the middle of the Japanese home islands. Your mission will be to break up enemy carrier striking forces by concentrating all power in suicide air attacks.”

Since I had never believed in the soundness of such tactics, the words “concentrating all power in suicide air attacks” were especially disagreeable to me. Accordingly, I frankly expressed my opinions concerning air operations and asked permission to decline this post which would call for carrying out assignments in which I did not believe. Admiral Ito went to consult with his Chief and returned shortly to say that I was still requested to take the post. Thereupon I accepted with the understanding that implementation of the attacks should be left to the decision of the Fifth Air Fleet commander. I insisted on this condition to insure that the Fifth Air Fleet could determine how and when each of its units would be employed. I knew that two of the eight units were practically untrained and so were not fit for anything but suicide duty. The others, capable of any kind of operation, should be kept flexible as to their employment as long as possible. The airfields of northern Kyushu and Shikoku were designated for training or rear area purposes, while the ones in southern Kyushu were to be operational bases.

Our new headquarters were established at Kanoya on February 13, and the direction of operations was begun. The headquarters office was located in a new barrack building set unobtrusively beside the airfield. This temporary structure was not even adequate to keep out the cold winter wind, and charcoal braziers were our only form of heating apparatus. But the approach of spring was soon heralded by white plum blossoms in the trees, delicate violets peeking from the weeds, and nightingales warbling in bushes outside the barrack windows.


On March 14 a reconnaissance plane from Truk reported the departure of an enemy task force from Ulithi. On March 17 at 2300 the alarm in the headquarters of the Fifth Air Fleet rang loudly to announce a report from one of our planes patrolling in waters to the south. It had made radar contact with a northbound fleet of enemy ships. Our planes, alerted since early afternoon, were ready and waiting to launch an attack on any enemy ships that might appear in our area of operations. This was the golden opportunity.

At 0350 next morning fifty torpedo planes took off from fields at Miyazaki and Kanoya, the night attack units followed directly by dawn attack units, in accordance with Battle Plan No. 1. And most of the night attack units succeeded in locating the enemy and launching their torpedoes.

Enemy planes in turn struck the Kanoya base at about 0500 and maintained incessant attacks for almost five hours, but these raids were so weak and disordered that it appeared we might have upset their assault plan. From dawn reconnaissance reports it was estimated that the enemy force was built around fifteen carriers, one of which had been set afire. Starting at 0600 our fighters and bombers took advantage of every interval between enemy attacks on the field to get into the air and assault the enemy ships.

From a hill-top vantage point I observed that most of the enemy attacks concen­trated on empty hangar and barrack buildings from which our colors waved enticingly in the breeze. There was no real damage done at Kanoya during the first day except for a few men injured by bomb fragments, but the battle raged day and night.

Reconnaissance planes sent out in the morning of the 19th reported that the enemy had only eleven carriers. This was encouraging, but communication wires at our land bases had now been severed by the bombardment, and radio channels were so crowded and confused that planes returning from combat were landing everywhere on Kyushu, Shikoku, and Chugoku. It was impossible for commanders to know how many planes they had or where they were, and direction of operations became extremely difficult.

On the 19th there were no enemy attacks in southern Kyushu, and at 1030 next morning the enemy task force was found to be withdrawing to the east. An immediate order for all units to pursue brought out more planes than I had counted on to continue the attack, and the day’s reports indicated that three more enemy carriers had been bombed.3 Following night and dawn sorties by our planes on the 21st, the enemy was spotted about 300 miles southeast of Kanoya. In an effort to take advantage of this opportunity eighteen bombers were ordered to take off at 1330, each bomber loaded with a 4,700-­pound Oka (cherry blossom) bomb.

This cumbersome cargo deprived a bomber of all maneuverability and made it a sitting duck for enemy fighter planes. Hence these bombers should have had plenty of escort planes for such a mission, but the fatigue after several successive days of battle left only thirty fighters to accompany the bombers. This force approached to within some sixty miles from the enemy fleet when it was intercepted and engaged by about fifty Grumman fighters. In ten minutes of combat the entire Japanese force was destroyed.

In three days of battle the Fifth Air Fleet lost 161 planes. Aerial reconnaissance indicated that five enemy carriers had been withdrawn from the battle line because of damage. These results were less than had been anticipated, but I was satisfied that the fighting flexibility of the Fifth Air Fleet had been maintained.


It did not seem that the enemy had even had time enough to regroup his fleet for another offensive, when, on March 23, raids were made on Okinawa and Minami Daito Islands, followed by naval shellings the next day. On the 25th the enemy landed on the nearby Kerama Islands as a preliminary to invading Okinawa. Having earlier placed the Third and Tenth Air Fleets under the Fifth Air Fleet, the Commander in Chief Combined Fleet now ordered that the combined force attack enemy forces around Okinawa. Our combined naval air strength amounted to about 1,815 planes, including 540 special attack planes, and the Sixth Army Air Force had been ordered to cooperate with us. Preparations were being made for this all-out effort when the enemy, on the first day of April, made their first landings on Okinawa. It had been settled with the Army that their planes would attack transports while our Navy planes would be responsible for the warships. Thus, target allocations had been made, but the only planes and pilots capable of attacking anything stronger than a trans­port were those of the original Fifth Air Fleet and a few elements of the Third.

Full preparations were completed and our first general attack was launched on April 6. In the early morning reconnaissance planes had spotted four U. S. task forces in the water south of Amami Oshima. From surrounding air bases we sent thirty dive bombers, forty fighter bombers, and fifty fighters to lassault “this enemy.” The only indication of results came from the intelligence section at Kanoya which judged, from enemy radio interceptions, that at least four carriers had been hit by bombs. Another thirty fighters from Kanoya took off at noon and flew directly to Okinawa. Their approach was coordinated with planes from other bases, all timed to catch the enemy fighters just after they had landed. A total of about a hundred of our fighters charged upon Okinawa in this effort to gain control of the air. Paced with their approach, three reconnaissance planes scattered counter-radar “window” to the east of Okinawa, successfully deceiving the enemy who took Minami Daito Island to be our main base and attacked it with most of his force. Seizing this opportunity some 110 naval suicide planes and ninety Army planes pulled a surprise attack from the west on U. S. warships and transports in the vicinity of Okinawa. The First Air Fleet and the Eighth Air Division from Formosa also contributed to this assault, which set fire to many ships and so filled the sky with smoke that reconnaissance planes could not see enough to make a clear report of the situation. The 32nd Army, which was defending Okinawa, reported that more than thirty ships were observed sinking, and more than twenty burning.

In the early morning of the 7th a scout plane found three carriers moving southeast­ward at slow speed and trailing oil. Of thirty planes dispatched promptly to attack them, 19 were reported to have plunged into the carriers. This same day saw seventy planes crash into enemy transports. And again in­tercepted enemy radio messages indicated that a great deal of damage had been done.

This day saw another form of suicide venture by the Japanese Navy that failed completely. Battleship Yamato was sunk, along with the light cruiser and three destroyers of her nine escorting warships, by enemy air attack. This force had sortied from the homeland the day before, heading for Okinawa. Its mission was to approach the enemy­besieged island and shell the landing areas with Yamato’s long-range, 18.1-inch guns. It was felt that the massed fire power of these ten ships could wreak profitable damage on the enemy aircraft they were bound to attract, and that at the same time Japanese suicide planes would have increased success in striking their targets without aerial competition, as a result of the American fighter protection being lured to the Yamato. Japanese resignation to the sacrifice of Yamato in this mission was apparent from the fact that on her departure from Japan she was provided with only enough fuel for a one-way run to Okinawa. As it turned out this was more than enough, for after two hours of intense air attack she sank at 1430 of the day following her sortie, and a full day before reaching her bombardment position off Okinawa.

However, we at Fifth Air Fleet headquarters received an encouraging message on the 9th, as follows: “In view of the impact your heavy assaults are having on the enemy, the Naval General Staff expects you to continue general attacks at all cost.” Accordingly, a second general attack consisting of a hundred suicide planes and 150 fighters was launched on April 12. On the same day another detachment of twenty planes attacked an enemy task force cruising in the waters east of Okinawa, and that night 45 torpedo planes made still another attack in which five warships were reported set afire.

A report from our naval attache in Portugal announced the serious losses sustained by the United States Navy and indicated that if the rate of attrition continued, the operation would result in disaster for the enemy. This inspired a third general attack by 220 Army and Navy planes on the night of April 15-16, while a separate detachment of 110 planes assaulted enemy task groups southeast of Kikai Island.

By this time it was obvious that our operations were making some progress, but, as might be expected of such tactics, they were also taking a terrific toll of our own air strength. It was estimated that 2,000 naval planes were involved in combat operations between March 23 and April 16, and some 600 of these were destroyed. As aircraft complements became depleted and replacements were not immediately available we were compelled to suspend operations for a short time. While the Navy took the attitude that these were last-ditch operations and devoted its entire energy to them, the Army regarded the Okinawa campaign as a mere preliminary to decisive battles in Japan proper, and so kept most of its strength in reserve. There was little, if any, unity of idea between the two services on this point.

On April 17 the Tenth Air Fleet was withdrawn from our command, leaving us only 610 planes, of which a mere 370 were operational. In addition our forces had been fighting day and night and everyone was beginning to show signs of fatigue. Under such circumstances, despite the urgency and our own eagerness, we were unable to launch continuous and successive attacks as before, but had to be content with sporadic air actions. A fourth general attack was finally launched on the 28th, but fewer than sixty planes could be mustered for the event.

Meanwhile the Army’s situation on Okinawa was in serious jeopardy under the enemy’s overwhelming fire power. The fresh green of the mountains north of Shuri had been transformed into a scorched brown. The 32nd Army’s elaborately worked out positions were reduced one after the other as the enemy moved steadily southward. This irritated the Army command into pressing a courageous and determined offensive on May 4 in which the Navy joined by carrying out a fifth general attack with 280 planes. As a result we were pleased to receive a message of thanks from the 32nd Army which reported that the offensive had made good progress. But our joy was fleeting, as the Army called off the offensive after suffering almost 7,000 casualties.

Our sixth general attack was executed on May 11 and the 240 participating planes achieved some results. A night patrol plane spotted an approaching enemy task force at 0300 on the 14th, but its report was delayed owing to a communication failure, and no planes were sent to intercept. From early morning until mid-afternoon, therefore, the air bases of southern Kyushu were raided by about 450 planes, but again our casualties were insignificant. With great patience our air units bided their time throughout the day and made preparations for a night attack. In our midnight attack a carrier was set on fire, but a sudden rain prevented our following up this opportunity.

On the night of May 24 a seventh general attack was made in cooperation with the Army, who had decided to make air attacks on Allied airfields on Okinawa. By this time we had been forced to include Shiragiku (white chrysanthemum) units in our operations. These units were made up of old training planes whose combat capabilities were practically nil, but they were used anyhow to fill the many gaps in our decimated organization. I had to smile, therefore, on reading an intercepted message from an enemy destroyer, “We are being pursued by a Jap plane making 85 knots.”

Two final efforts were made on May 27 and June 7 when we executed our eighth and ninth general attacks, but so few planes were available each time that we were obliged to acknowledge the failure of these attempts. The 32nd Army continued its reluctant southward retreat in the latter part of May, withdrawing from the Shuri redoubt to the Kibuya peninsula at the southern tip of Okinawa. It was greatly hampered by the seasonal rains, which also interrupted our feeble efforts to help. Rejecting an enemy order to surrender, the Commander, Lieutenant General Ushizima, and his Chief of Staff committed suicide at their command post on June 23, and eighty days of fierce battle on Okinawa came to an end.

In these operations, in which a total of 3,000 planes took part and more than 700 were lost, the Japanese Navy utterly exhausted itself. The battle for Okinawa proved conclusively the defects of suicide air attacks. Such operations cannot be successful where materiel and trained manpower is limited. It would have been far wiser for the sadly depleted Japanese military to have conserved its manpower instead of squander­ing it as was done. It is not strange that this unrealistic aerial tactic ended in failure. Even the physical destructive power of the weapon itself was not sufficient for the task for which it had been designed. While it might deal a fatal blow to small warships or transports, the enemy aircraft carriers, which were meant to be primary targets, were sometimes able to survive attacks in which they were hit several times. Setting aside Admiral Ohnishi’s original concept of adopting suicide attacks for the limited purpose of inactivating carrier decks for a week, the whole concept of suicide attacks to annihilate enemy task forces was more than unreasonable, it was sheer lunacy. Once the order had been issued by Headquarters for these suicide attacks, they lost their volunteer aspect and became, instead, “murder attacks,” and humanity was lost sight of.

As might have been expected, these attacks created many command problems. Early in the Okinawa campaign pilots could go to their death with some hope that their country might realize some benefit from their sacrifice. But toward the last, the doomed pilots had good reason for doubting the validity of the cause in which they were told to die. The difficulties became especially apparent when men in aviation training were peremptorily ordered to the front and to death. When it came time for their take off, the pilots’ attitudes ranged from the despair of sheep headed for slaughter to open expressions of contempt for their superior officers. There were frequent and obvious cases of pilots returning from sorties claiming that they could not locate any enemy ships, and one pilot even strafed his commanding officer’s quarters as he took off. When planes did not return there was seldom any way of knowing the results of their sacrifice. There was no conclusive means of determining if they had crashed into a target. Even with reconnaissance planes accompanying, there was no assurance of a valid report because observers were usually kept too busy avoiding enemy fighters. After a careful examination of all reports I estimated that only a total of five enemy carriers had been damaged by Kyushu-based planes. Yet a single unit commander gave me absolute assurance that his men alone had damaged eight carriers, and he continued, “If the results achieved are going to be so underestimated, there is no justification for the deaths of my men. If Headquarters will not acknowledge these achievements at full value I must commit harakiri as an expression of my disapproval and by way of apology.”


Upon conclusion of the Okinawa campaign the Japanese Naval General Staff produced an estimate of the situation which summed up prospective moves of the enemy.

First, Allied forces would attempt to seize the southwestern home islands of Tokunoshima, Kikaigashima, and Amami Oshima. At the same time his air forces at Okinawa, Iwo Jima, and in the Marianas would increase their bombing operations against Japan so as to lower morale, destroy airfields, and disrupt land and sea communications.

Second, just before invading the homeland, Allied task forces would concentrate on smashing the Japanese fleet and air bases. The most likely points of invasion are in southern Kyushu and in the Kanto Plain area which surrounds Tokyo, but there is a strong probability that landings may also be made at Saishuto, south of Korea, in order to sever communications between Korea, Manchuria, and Japan. Landings in southern Kyushu would take place after August and would be attempted by 15-20 divisions. When a foothold had been established there, an invasion of the Kanto Plain by some 30­40 divisions would follow, some time after September.

A further breakdown of enemy capabilities and intentions was made with predictions that one division would hit the Koshiki Islands, off Kyushu, to set up a base; no fewer than six divisions would assault the western coast, north of Cape Noma; and when a bridgehead was established there the main force of at least eight divisions would land at Ariake Bay, on the eastern coast of Miyazaki. In the east a minimum of eight divisions would lead off by striking at Oshima, Kujukurihama, and Kashimanada, followed by about twelve divisions on the Bozo Peninsula, and then the main force would land at Sagami Bay and sweep over the Kanto Plain.

It was indicated that the Allies would have the following naval forces available to carry out these operations:

Okinawa Campaign002
To oppose these forces, in August, 1945, Japan’s principal strength lay in its four Naval Air Fleets which contained a total of 5,044 planes, in categories as follow:

Okinawa Campaign003

Of ships and naval craft Japan had the following:

Okinawa Campaign004


Okinawa Campaign005
Statistics of Pocket Submarines and Manned Torpedoes

For the benefit of readers unfamiliar with craft of the last three categories, salient specifications of each are given.

The charge-laden motorboats (Shinyo) were 16 1⁄2 feet in length, weighed 2,800 pounds, had a top speed of 23 knots, a cruising radius of 250 miles, and carried a 550-pound explosive charge in the bow.

A careful study of the situation led the Naval General Staff to conclude that all Japanese surface forces should be meticulously conserved until the enemy actually began landing. To this end all warships were ordered into concealment at various Inland Sea islands north of Iyonada. And, following the pattern I had established earlier in southern Kyushu, a system of airfield defense was instituted throughout the country.

It was determined that special attack units, both air and submarine, should seize every opportunity to hit enemy ships at advanced bases before their task forces sortied. Careful and thorough patrols were flown day and night by some 140 planes which covered the sea to a distance of 600 miles from each of the anticipated landing points. Coastal waters were patrolled by small and medium submarines.

As soon as the enemy’s intended landing points were indicated, almost all Japanese air and sea forces would immediately be concentrated in the battle area. There were 150 specially equipped torpedo planes set up in night attack groups to strike warships bombarding our coast, and 330 top-notch planes and pilots constituted elite units which were assigned to engage approaching enemy task forces. Another group of planes-a hundred transports-were designated to carry 1,200 airborne troops to Okinawa airfields where they would destroy planes and fuel dumps at the critical moment. Most of the air forces were disposed in the southwestern part of the homeland, while the surface and underwater suicide units were distributed about equally between the eastern and western areas as follows:

Okinawa Campaign006
As enemy convoys finally approached their assault destinations, all suicide forces were to launch day and night attacks which would be sustained for at least ten days. It was expected that 3,725 naval planes (more than half of them trainers), and 2,500 army planes would be available to carry out these attacks. At the same time a squadron composed of one cruiser and nineteen destroyers was to make night attacks on the ships once they had anchored. It was calculated that these combined efforts would succeed in sinking half of the anticipated ships in the enemy invasion fleet.

But this whole scheme was conceived in futility and prepared for in despair. So critical was the situation by August, 1945, that Japan’s shortage of fuel alone was enough to banish confidence. The contemplated ten­day aerial effort would have drained every drop of fuel from every plane and storage tank in Japan. The only commodity to survive the pinch of Japan’s wartime economy was hope, and that was rapidly disappearing.


An official announcement by Combined Fleet Headquarters said that 2,409 Kamikaze pilots were killed in performance of their duty during World War II. It was a real scourge of Japan’s military forces that permitted human life to be treated so lightly through a misinterpretation of the true spirit of Bushido.

Japan’s suicide air operations mark the Pacific War with two scars that will remain forever in the annals of battle: one, of shame at the mistaken way of command; the other, one of valor at the self-sacrificing spirit of young men who died for their beloved country.