U.S. Department of the Army No Gun Ri Review Report/Executive Summary

EXECUTIVE SUMMARY

Introduction

Following the release of the Associated Press story concerning the matter on September 29, 1999, the United States (U.S.) and the Republic of Korea (ROK) initiated independent, but cooperative, reviews of the incident at No Gun Ri. This story brought to the forefront the earlier efforts of Korean citizens to secure an official inquiry into their claims surrounding certain events that occurred in the vicinity of No Gun Ri, including the firing upon Korean refugees at the double railroad overpass and an air strike on the railroad track.

Over the last year, the U.S. Review Team has conducted an exhaustive factual review by examining over a million documents from the National Archives, conducting interviews with approximately 200 American witnesses, and analyzing the interview transcripts and oral statements of approximately 75 Korean witnesses. The U.S. Review Team also closely examined press reports, aerial imagery, and other forensic examination results. This U.S. Report reflects the U.S. Review Team’s factual findings based upon all the evidence available on the incident.

Unfortunately, the passage of 50 years greatly reduces the possibility that we will ever know all of the facts surrounding this particular event. A large number of factors, including but not limited to trauma, age, and the media, influenced the recollection of Korean and U.S. witnesses. By comparing and contrasting all of these available information sources, the U.S. Review Team has developed a clearer picture of the events that occurred in the vicinity of No Gun Ri in July 1950. The findings of the U.S. Review Team have been organized into several key issues, which describe the Team's conclusions regarding what occurred at No Gun Ri based upon all the information available half a century later.

I. Background - The Korean Account

The Korean villagers stated that on July 25, 1950, U.S. soldiers evacuated approximately 500 to 600 villagers from their homes in Im Gae Ri and Joo Gok Ri. The villagers said the U.S. soldiers escorted them towards the south. Later that evening, the American soldiers led the villagers near a riverbank at Ha Ga Ri and ordered them to stay there that night. During the night, the villagers witnessed a long parade of U.S. troops and vehicles moving towards Pusan.

On the morning of July 26, 1950, the villagers continued south along the SeoulPusan road. According to their statements, when the villagers reached the vicinity of No Gun Ri, U.S. soldiers stopped them at a roadblock and ordered the group onto the railroad tracks, where the soldiers searched them and their personal belongings. The Koreans state that, although the soldiers found no prohibited items (such as weapons or other military contraband), the soldiers ordered an air attack upon the villagers via radio communications with U.S. aircraft. Shortly afterwards, planes flew over and dropped bombs and fired machine guns, killing approximately 100 villagers on the railroad tracks. Those villagers who survived sought protection in a small culvert underneath the railroad tracks. The U.S. soldiers drove the villagers out of the culvert and into the larger double tunnels nearby (this report subsequently refers to these tunnels as the "double railroad overpass"). The Koreans state that the U.S. soldiers then fired into both ends of the tunnels over a period of four days (July 26–29, 1950), resulting in approximately 300 additional deaths.

II. Department of Defense Review Directives

On September 30, 1999, the Secretary of Defense directed the Secretary of the Army to lead a review to determine "the full scope of the facts surrounding these [No Gun Ri] press reports." On October 25, 1999, the Secretary of the Army directed The Inspector General to conduct a thorough review of the allegations, pursue every reasonable lead to determine the facts, and then prepare and submit a report of the findings with regard to the allegations.

The Office of the Secretary of Defense established a Steering Group chaired by the Under Secretary of Defense (Personnel and Readiness) to oversee the conduct of the review. In addition, the Secretary of Defense invited eight distinguished Americans, who are not affiliated with the Department of Defense, to advise on the conduct of the review based upon their expertise in academia, journalism, the Korean War, and U.S.ROK relations.

III. Department of the Army Inspector General Review Effort

The Inspector General developed a four-phase concept plan: Preparation; Research and Interviews; Review and Analysis; and Production of the Final Report. The Inspector General then formed the No Gun Ri Review Team (U.S. Review Team) into a Research Team and an Interview Team. The research effort, led by an Army historian, began in October 1999. The Research Team consisted of Department of the Army military and civilian members augmented by a United States Air Force research team, an imagery analyst, a Korean linguist, and professional research assistants from the U.S. Army Corps of Engineers. The researchers examined over one million pages of text from the National Archives and other repositories and approximately 45,000 containers of United States Air Force reconnaissance film.

The interview process started on December 29, 1999, after the Interview Team located former soldiers assigned to the major combat units that passed through the Yongdong-Hwanggan area in mid- to late July 1950. The Interview Team and Air Force researchers culled through over 7,375 names to locate and interview approximately 200 U.S. veterans. While every effort was made to make this a comprehensive sample, the U.S. Review Team had no power to compel a witness to grant an interview and no authority to issue subpoenas or to grant immunity. In fact, eleven veterans contacted by the U.S. Review Team declined to be interviewed. The U.S. Review Team did review, however, the published accounts of some witnesses who declined to be interviewed by the Team.

IV. U.S. and ROK Cooperation

The Department of the Army and the Department of Defense worked in close cooperation with the representatives of the government of the Republic of Korea who were conducting a parallel review of the allegations. Members from the U.S. Review Team, the Republic of Korea Investigation Team (ROK Review Team), and government officials from both countries met on approximately a dozen occasions in both the United States and Korea, to include the Secretary of the Army’s meetings with President Kim Dae-Jung and Minister of National Defense Cho Song-Tae in January 2000. The U.S. Review Team provided the ROK Review Team with copies of all relevant documents and other information discovered in the course of the review in support of the ROK's parallel investigation. On two occasions, the U.S. Review Team supported working visits by a ROK Review Team researcher to the National Archives in College Park, Maryland. The U.S. Review Team provided full access to, and funded the reproduction costs of, any materials already gathered by the U.S. researchers. No information was withheld.

V. Organization of the U.S. Report

The U.S. Review Team conducted this review and prepared this report fully aware of the political, military, and emotional significance of the allegations. This report is not intended as a point-by-point response to the media and Korean accounts. The report presents an independent assessment of the facts derived directly from an exhaustive review of primary and secondary sources, the statements of U.S. veterans and Koreans, ballistic and pathology forensics, and imagery analysis.

The report consists of an Executive Summary, five chapters, and five appendices. Chapter 1 (Introduction) outlines the purpose, background, and overall organization and conduct of the review. Chapter 2 (Background and History) describes the ground events unfolding on the Korean Peninsula in July 1950. Chapter 3 (Combat Operations—July 1950) examines the state of U.S. intelligence and U.S. ground forces in July 1950 and provides a day-by-day account of the tactical operations of the 1st Cavalry Division in the vicinity of No Gun Ri during the last week of July 1950. This chapter also includes research on U.S. and allied air operations in the Yongdong-Hwanggan area for the same time period. Several photographs from 1950 are inserted between Chapters 2 and 3. Chapter 4 (Analysis of Interview Data) provides the analysis of interviews of American and Korean witnesses. The review of witness statements identifies areas of consensus between statements and outlines possible sequences of events. Finally, Chapter 5 (Key Issue Analysis and Findings) synthesizes the analysis of documentary research and witness interviews into a thorough, fact-based set of findings.

The appendices supplement the material in the main body of the report. Appendix A (Research Methodology) documents in detail the methodology used in the research of the historical records. Appendix B (Forensic Evidence) provides an analysis of the forensic evidence associated with the No Gun Ri site. This appendix discusses the sources of Korean casualty estimates, analysis of the ballistic evidence collected by Korean authorities, and an analysis of the USAF reconnaissance film taken over the No Gun Ri area on August 6, 1950. Appendix C (Imagery Analysis) contains the analysis of the August 6, 1950, USAF reconnaissance photograph performed by the National Imagery and Mapping Agency (NIMA). This appendix includes the NIMA response to the ROK Investigation Team's questions concerning this analysis. Appendix D (Joint Cooperation) discusses the actions taken to ensure a cooperative and coordinated effort between the ROK and U.S. Review Teams, including joint meetings and the exchange of documents and other information. Appendix E (Supporting Documents) contains explanatory charts and maps.

VI. Findings

Given the challenge of ascertaining facts a half century after their occurrence, the U.S. Review Team made findings when possible, identified possibilities, and noted when the evidence was not sufficient to identify a possibility or reach a finding about what may have occurred at No Gun Ri in July 1950 based upon an analysis of available information. A summary of its factual findings has been organized into several key issues. These issues were identified and developed in coordination with the Office of the Secretary of Defense Steering Group, U.S. Outside Experts, and counterparts from the Republic of Korea.

 

A. Key Issue 1: Condition of U.S. Forces in July 1950

Background. U.S. soldiers were young, under-trained, under-equipped, and unprepared for the fight they would wage against the North Korean People's Army (NKPA). The soldiers of the Army of Occupation in Japan functioned primarily as a constabulary in a conquered land and not as combat-ready warfighters.

Their lack of combat preparedness was a direct result of deficiencies in training, equipment, structure, personnel strength, and leadership. Proper training areas were not available to conduct more than small-unit training. Classes for critical specialties such as maintenance and communications were also inadequate. Most of their equipment, including ammunition, was of World War II vintage, and had been poorly stored and maintained. The three infantry regiments in the 1st Cavalry Division had only two of the three battalions normally assigned. Likewise, each regiment lacked its authorized tank company, and the division artillery battalions contained only two of the normal three firing batteries. In response to a requirement to bring the 24th Infantry Division up to strength prior to that division's departure for Korea, the 1st Cavalry Division transferred nearly 800 men, most of them from the top four senior non-commissioned officer grades, to the 24th. This loss of non-commissioned officers with whom the soldiers had trained weakened the cohesion of the division and significantly reduced the number of leaders with combat experience at the small-unit level.

Finding. Based on the documentary evidence, as well as the statements by U.S. veterans, the U.S. Review Team concluded that most American units and soldiers were not adequately prepared for the combat conditions that they confronted in Korea in June and July 1950. No experience or training equipped them to deal with an aggressive enemy that employed both conventional and guerilla warfare tactics or with a large refugee population, which the enemy was known to have infiltrated. Shortages of experienced Non-commissioned officers, along with inadequate equipment and doctrine, made it difficult for individuals or units to adapt to these conditions.

 

B. Key Issue 2: U.S. and ROK Refugee Control Policies

Background. The U.S. troops were completely unprepared for the stark reality of dealing with the numerous, uncontrolled refugees who clogged the roads and complicated the battlefield to an unexpected degree. Early on in the war, U.S. forces encountered the NKPA practice of infiltrating soldiers dressed as civilians among large refugee concentrations. Once behind American lines, these infiltrated soldiers would then conduct guerilla-style combat operations against American rear-area units and activities.

In late July 1950, the ROK government and the Eighth U.S. Army Headquarters issued refugee control policies to protect the U.S. and ROK forces from NKPA infiltration and attacks from the rear. Additionally, these policies were aimed at reducing the adverse impact of large refugee concentrations on main supply routes, which stymied the U.S. and UN troops' ability to rush ammunition forward and evacuate casualties to the rear. These U.S. and ROK refugee policies depended heavily upon the constant presence of, and coordination with, the ROK National Police to handle the uncontrolled refugee population.

Despite comments attributed to Major General Gay, the 1st Cavalry Division Commander, that he would not employ the Korean National Police in his division's area of operations, his refugee policy directive of July 23, 1950, made the National Police responsible for handling refugees. The movement of civilians and refugees in the 1st Cavalry Division area was restricted to specific hours and for specific purposes by a limited number of people, and the National Police were responsible for enforcing the policy.

On July 26, 1950, the Eighth U.S. Army Korea (EUSAK), in coordination with the ROK government, established and disseminated a plan to control refugee movement which:

- precluded movement of refugees across battle lines at all times, prohibited evacuation of villages without general officer approval, and established a National Police responsibility,
- prescribed procedures for Korean National Police to clear desired areas and routes,
- strictly precluded Korean civilian movement during the hours of darkness, and
- established requirements for disseminating the policy.

The Eighth Army's policy was intended to deny the NKPA their widely used infiltration tactic while also safeguarding civilians by prohibiting refugees from crossing battle lines (battle lines are the areas where there is contact with the enemy or contact is about to occur). The policy did not state that refugees could not cross friendly lines and contains instructions for the handling of refugees in friendly areas (friendly lines are the forward troop positions not in contact with the enemy). The policy emphasized the Korean government's responsibility for the control and screening of refugees to provide for their welfare. Nothing in this policy was intended to put refugees at risk.

Most veterans from the 7th Cavalry Regiment interviewed by the U.S. Review Team were enlisted men during the Korean War and did not receive copies of policies from higher headquarters. In general, most U.S. veterans remembered warnings that there were North Korean infiltrators among the refugees. The veterans who remembered more specific details about refugee control remembered specific actions to be taken; for example, keep refugees off the roads, do not let refugees pass, or search refugees and let them pass.

Finding. From its study of the refugee control policies in effect during the last week of July 1950, the U.S. Review Team found that the Eighth U.S. Army published, in coordination with the ROK government, refugee control policies that reflected two predominant concerns: (1) protecting U.S. and ROK troops from the danger of NKPA soldiers infiltrating U.S.–ROK lines; and (2) precluding uncontrolled refugee movements from impeding flows of supplies and troops. The published 1st Cavalry Division refugee control policy dated July 23, 1950, reflected the same two concerns. The task of keeping innocent civilians out of harm's way was left entirely to ROK authorities. By implication, these policies also protected refugees by attempting to ensure they were not in harm's way.

 

C. Key Issue 3: Tactical Situation July 22–29, 1950

Background. The 1st Cavalry Division relieved the 24th Infantry Division northwest of Yongdong on July 22, 1950. The 7th Cavalry Regiment of the 1st Cavalry Division arrived in Pohangdong, Korea, on July 22, 1950, and the 2nd Battalion, 7th Cavalry Regiment, moved forward to the Yongdong area. With friendly forces outnumbered by the NKPA, the Eighth Army developed a strategy to withdraw behind the last defensible terrain feature, the Naktong River. As events developed, the 1st Cavalry Division withdrew from Yongdong through a series of delaying actions in accordance with the Eighth Army strategy and to avoid a threatened envelopment. On the evening of July 25, 1950, the 7th Cavalry Regiment was supporting the 5th Cavalry Regiment in positions east of Yongdong.

Sometime during the night of July 25, the 7th Cavalry received a report that a breakthrough had occurred in the sector to the 7th Cavalry Regiment's north.

Finding. The U.S. Review Team found that, in the early morning hours of July 26, 1950, the 2nd Battalion of the 7th Cavalry Regiment, without specific orders but believing they were being enveloped, conducted a disorganized and undisciplined withdrawal from a position east of Yongdong to the vicinity of No Gun Ri. They spent the remaining hours of July 26 until late into that night recovering abandoned personnel and equipment from the area where the air strike and machine-gun firing on Korean refugees is alleged to have occurred. On July 26, 1950, at 9:30 at night, 119 men were still unaccounted for. It will probably never be possible to reconstruct the activities of the scattered soldiers of the 2nd Battalion.

The U.S. Review Team determined that the 1st Battalion, 7th Cavalry Regiment, arrived in the vicinity of No Gun Ri in the afternoon of July 26, 1950. They relieved the 2nd Battalion, 5th Cavalry Regiment, and established their position east of the 2nd Battalion, 7th Cavalry Regiment.

The U.S. Review Team found that there was repeated contact reported between the 7th Cavalry and enemy forces in the vicinity of No Gun Ri on July 27 and July 28. The records indicate by this time that the 7th Cavalry had been told that there were no friendly forces to the west and south of No Gun Ri (i.e. back toward Yongdong). The 2nd Battalion, 7th Cavalry, reported an enemy column on the railroad tracks on July 27, which they fired upon. On July 29, the battalion withdrew as the NKPA advanced.

The U.S. Review Team concluded that based on the available evidence, the 7th Cavalry Regiment was under attack, as they believed, between July 27 and July 29, 1950, when in position near No Gun Ri.

 

D. Key Issue 4: Assembly and Movement of Villagers

Background. The U.S. and ROK policy in July 1950 stated generally that Korean civilians should not evacuate their villages. The U.S. Review Team could not determine the reasons why the refugees gathered in Im Gae Ri, but this gathering of refugees was probably not the result of any U.S. action. Some witnesses stated that the Americans told them that they were being moved for their safety. Some U.S. veterans remember escorting refugees from villages, but these veterans cannot remember the villages' names or the dates the evacuations occurred. Therefore, the U.S. Review Team cannot rule out the possibility that U.S. soldiers told the villagers at Im Gae Ri to evacuate the village.

While the U.S. Review Team cannot rule out the possibility that the villagers were moved, there was no sound military reason for soldiers to travel approximately three miles off their designated movement route to the village of Im Gae Ri during a hasty withdrawal for the purpose of encouraging an additional 400 refugees onto the already crowded roads and further aggravating the congested conditions. It is also unlikely that the soldiers would have performed this evacuation given the widespread knowledge and fear of North Korean infiltrators believed to be present in refugee concentrations.

Some 7th Cavalry Regiment veterans recalled displacing South Koreans from unknown villages on unknown dates. The U.S. Review Team found that the 7th Cavalry Regiment was not in the vicinity of Im Gae Ri on July 25 based upon official records of the Regiment's positions. Some veterans of the 5th Cavalry Regiment indicated that they evacuated or escorted Korean civilians from unknown villages in late July and early August 1950. A patrol from the 5th Cavalry Regiment may have told the villagers who had assembled at Im Gae Ri to leave.

Finding. The U.S. Review Team could not determine the reasons why the refugees gathered in Im Gae Ri, but the U.S. Review Team concluded that this gathering of refugees was probably not the result of U.S. action. Based on some of the available evidence, the U.S. Review Team cannot rule out the possibility that U.S. soldiers told the villagers at Im Gae Ri to evacuate the village, but the soldiers who did so were not from the 7th Cavalry Regiment.

 

E. Key Issue 5: Air Strikes in the Vicinity of No Gun Ri

Background. Korean witnesses describe an air strike / strafing around noon on July 26, 1950 on the railroad tracks. The Korean witnesses describe the effects of machine gun fire and explosions.

The U.S. Review Team could not locate any records to substantiate the occurrence of an air strike / strafing incident in the vicinity of No Gun Ri around noon on July 26, 1950. While there are mission reports for July 26, 1950, that could not be located, the missions can be accounted for through other reports. The only documented USAF air strike in the immediate vicinity of the Hwanggan area occurred southwest of No Gun Ri on July 27. This air strike was a friendly fire incident in which a F-80 accidentally strafed the 1st Battalion, 7th Cavalry Regiment's command post at 7:15 in the morning. The strafing destroyed two U.S. trucks but claimed no lives.

The friendly strike on the 1st Battalion, 7th Cavalry Regiment, caused the 7th Cavalry Regimental Commander to request immediately that he be assigned a Tactical Air Control Party in order to control aircraft in his area and thereby preclude further friendly-fire incidents. Only a Tactical Air Control Party (TACP) with a jeep-mounted AN/VRC-1 radio could talk to the Air Force elements, including the strike aircraft. There was only one TACP operating in support of the 1st Cavalry Division during this period of time. This TACP was not located in the vicinity of No Gun Ri during the period of July 26 to July 29, 1950. Ordinary soldiers could not communicate on their radios with aircraft. Although it was possible for the Army to request an air strike from the Air Force, the process was cumbersome and took considerable time because the request had to be processed through Army and Air Force channels.

No U.S. Air Force veteran that the U.S. Review Team interviewed participated in, or had any knowledge of anyone participating in, the strafing of civilians in the vicinity of No Gun Ri in late July 1950. U.S. Air Force interviewees vividly recalled stern verbal policies implemented to prevent the attack of non-combatants.

The Navy discovered no evidence of naval aircraft operating in the vicinity of No Gun Ri on July 26 or 27. However, on July 28, Navy aircraft from the USS Valley Forge were directed into the area and attacked a railroad tunnel occupied by enemy troops and other targets forward of the 7th Cavalry in the direction of Yongdong with bombs and machine guns.

The Defense Intelligence Agency found 8th Tactical Reconnaissance Squadron photographs of the No Gun Ri area dated August 6 and September 19, 1950. The Air Force Team showed these photographs to four retired photo interpreters of national reputation, all of whom agree that the photographs show no signs of bombing or strafing on the railroad tracks. A NIMA photo interpreter maintains that some patterns near the tracks approximately 350 yards from the double railroad overpass show "an imagery signature of probable strafing" but no bomb damage. The location of the probable strafing is in the same relative location identified by the Korean witnesses as that location where they were strafed.

Finding. An exhaustive search of U.S. Air Force and U.S. Navy records and interviews with U.S. pilots did not identify an air strike in the No Gun Ri area on July 26, 1950. The number of Korean witness statements describing the strafing and the photograph interpretation by NIMA does not permit the U.S. Review Team to exclude the possibility that U.S. or allied aircraft might have hit civilian refugees in the vicinity of No Gun Ri during an air strike / strafing on July 26, 1950. On July 27, 1950, an air strike did in fact occur on the 1st Battalion, 7th Cavalry's position near No Gun Ri that both the Air Force and Army recorded in official documents. On July 28, there was also an air strike on NKPA forces near 1st Battalion, 7th Cavalry Regiment. Assuming Korean civilians were near the positions of these strikes, they could have been injured.

The U.S. Review Team concluded that strafing may have occurred near No Gun Ri in the last week of July 1950 and could have injured or killed Korean civilians but that any such air strikes were not deliberate attacks on Korean civilians. The U. S. Review Team concluded that any air strikes / strafing occurring on July 26 took place under the same conditions as the air strikes / strafing on July 27, specifically an accidental air strike / strafing caused by the misidentification of targets and not a pre-planned strike. An accidental air strike / strafing could have happened due to several factors: target misidentification, lack of reliable communications, absence of a Tactical Air Control Party in the 7th Regiment, and the fluid nature of the battlefield. It was not a preplanned strike on civilian refugees.

 

F. Key Issue 6: Ground Fire in the Vicinity of No Gun Ri

Background. Some U.S. and Korean witness statements indicate that U.S. ground forces fired toward refugees in the vicinity of No Gun Ri during the period July 26–29, 1950, as discussed below. According to the Korean description of the events on July 26, 1950, refugees were strafed or bombed on the railroad tracks. Some fled the area or hid in ditches and others went into the double railroad overpass tunnel where they were fired upon from different locations for a period of up to four days, with the heaviest fire occurring on July 26 (which was the first day they report spending in the double railroad overpass).

In interviews, some U.S. veterans stated they saw or heard firing of various types including machine-gun, mortar, artillery, and rifle fire, near unidentified individuals in civilian clothing outside the tunnels / bridges in the vicinity of No Gun Ri. Only a few veterans interviewed by the U.S. Review Team stated they fired toward civilians in the vicinity of No Gun Ri. Two veterans fired over the heads of or into the ground in front to keep the civilians pinned down or to prevent them from moving. Several other veterans stated they either received hostile fire from, or saw hostile fire coming from, the civilian positions in the double railroad overpass and elsewhere. They also stated that they returned fire, or observed fire being returned, on the civilian positions as a response to the hostile fire they received or observed. Some veterans also remember intermittent NKPA and U.S. artillery and mortar fires.

Official records indicate that the NKPA attacked the 7th Cavalry on July 27 and 28, and the 7th Cavalry employed every means at its disposal to defend itself, including the use of small-arms fire, mortars, and artillery.

Finding. Although the U.S. Review Team cannot determine what happened near No Gun Ri with certainty, it is clear, based upon all available evidence, that an unknown number of Korean civilians were killed or injured by the effects of small-arms fire, artillery and mortar fire, and strafing that preceded or coincided with the NKPA's advance and the withdrawal of U.S. forces in the vicinity of No Gun Ri during the last week of July 1950. These Korean deaths and injuries occurred at different locations in the vicinity of No Gun Ri and were not concentrated exclusively at the double railroad overpass.

Some U.S. veterans describe fire that lasted for a few to 60 minutes. Some Korean witnesses describe fire day and night on the tunnel for as long as four days. Because Korean estimates of the length of time they spent in the tunnel are so inconsistent, the U.S. Review Team drew no conclusion about the amount of time they spent in the tunnel.

The firing was a result of hostile fire seen or received from civilian positions or fire directed over their heads or near them to control their movement. The deaths and injuries of civilians, wherever they occurred, were an unfortunate tragedy inherent to war and not a deliberate killing.

 

G. Key Issue 7: Issuance of Orders to Fire on Refugees

Background. To determine if soldiers or pilots were issued orders to attack and fire on refugees in the vicinity of No Gun Ri, the Review Team reviewed documents and conducted interviews with Army and Air Force veterans. Based upon the available evidence, which included the statements of veterans, documents, and the absence of documents, the U.S. Review Team concluded that U.S. commanders did not issue oral or written orders to fire on refugees in the vicinity of No Gun Ri between July 25 and 29, 1950.

Pilots were not ordered to attack and kill civilian refugees in the vicinity of No Gun Ri. Air strikes in the vicinity of No Gun Ri on July 26 were either the result of a misidentification of a target or an accident as discussed above. No USAF veteran that the U.S. Review Team interviewed participated in, or had any knowledge of anyone participating in, the strafing of civilians in the vicinity of No Gun Ri in late July 1950. U.S. Air Force interviewees vividly recalled stern verbal policies implemented to prevent the attack of non-combatants. In interviews, pilots stated that they sought out targets such as tanks, trucks, moving troops, and groups of men in uniform. Pilots fired when they were told a target was hostile and fired back when fired upon.

The U.S. Review Team found two documents that refer to an unknown Army request to the Air Force and the Navy to strafe civilian or refugee columns. The first reference is in a memorandum by COL T.C. Rogers, Fifth Air Force ADVON (Korea), dated 25 July 1950. The second reference is a Naval Activity Summary for the same date from the Aircraft Carrier Valley Forge. The U.S. Review Team could not find any originating request from the Army that prompted these two references. The Rogers' memorandum actually recommends that civilians not be attacked unless they are definitely known to be North Korean soldiers or have committed hostile acts. The Navy document stated that the first pass over personnel would be a non-firing run to identify if civilians were present. If the target was determined to be hostile, a firing run would follow.

Soldiers were not ordered to attack and kill civilian refugees in the vicinity of No Gun Ri. The veterans interviewed said that deadly force was not authorized against civilian refugees who posed no threat to the unit, and they were not given orders to shoot and kill civilian refugees in the vicinity of No Gun Ri. However, the U.S. Review Team found that soldiers who were in the vicinity of No Gun Ri were given an order to stop civilians and not to let them pass their position. Some soldiers did believe if civilian refugees did not stop, they could use deadly force to prevent them from passing.

Several other veterans stated they observed firing at the civilians in response to perceived hostile fire from the positions near the double railroad overpass and elsewhere. Based on veterans' interviews, the U.S. Review Team found that soldiers believed that they could take action in self-defense against civilians; that is, if they were fired upon or if they saw actions that indicated hostile intent. Some veterans said they observed firing in the direction of the double railroad overpass in response to fire from that location. Return fire in this case would have been an action in self-defense, and no orders were required. Two veterans fired over the heads of civilians, or into the ground in front to keep the civilians pinned down or prevent them from moving. The U.S. soldiers were repeatedly warned that North Korean soldiers wore civilian clothing over their uniforms in order to infiltrate U.S. positions. The U.S. soldiers were also told that North Korean soldiers would hide within refugee columns.

In interviews with the U.S. Review Team, several veterans stated that they assumed there was an order to fire on civilians because artillery and mortar fires were used that may have hit civilians. These veterans had no information to support their assertions. When interviewed, the veterans said they did not know who gave the order, they did not hear the order, they did not know when the order was given, and they personally did not receive the order. Former officers of the 2nd Battalion, 7th Cavalry Regiment, that the U.S. Review Team interviewed remain adamant that the battalion commander issued no order to fire on refugees at any time

There are references that appear to authorize firing on Korean civilians in Army records. The first reference was an abbreviated message that appeared in an 8th Cavalry Regiment message log dated 10:00 AM on July 24, 1950, that stated: "No refugees to cross the frontline. Fire everyone trying to cross the lines. Use discretion in case of women and children." The U.S. Review Team found no similar entry in the records of the 1st Cavalry Division, its other two regiments (the 5th and 7th Cavalry Regiments), or in the records of units subordinate to the 8th Cavalry Regiment. The U.S. Review Team found no evidence that the 8th Cavalry message was transmitted to the 5th or 7th Cavalry Regiments or any other subordinate element of the division. In interviews, U.S. veterans in the vicinity of No Gun Ri do not recall instructions to fire on civilian refugees. The 7th Cavalry Regiment was the unit in the vicinity of No Gun Ri on July 26. By July 26, 1950, the last elements of the 8th Cavalry Regiment were withdrawing from the vicinity of No Gun Ri to the division rear near Hwanggan.

The refugee control policy set by the 1st Cavalry Division Commander in his order of July 23, 1950, titled "Control of Refugee Movement" makes no mention of the use of force by soldiers. It stated: "Municipal authorities, local police and the National Police will enforce this directive." The U.S. Review Team concluded that the 8th Cavalry Regiment log entry did not constitute a 1st Cavalry Division order to fire upon Korean civilians at No Gun Ri.

The second reference was a 25th Infantry Division Commander's memorandum to commanders issued on July 27, 1950. On July 25, 1950, the 25th ID Activities Report stated: "Refugees and Korean Civilians were ordered out of the combat zone in order to eliminate possible serious traffic problems and to aid in blocking the infiltration of North Korean Forces through the lines. These instructions were passed to the civilians through the Korean Police." The July 27, 1950, memo to Commanders reads: "Korean police have been directed to remove all civilians from the area between the blue lines shown on the attached overlay and report the evacuation has been accomplished. All civilians seen in this area are to be considered as enemy and action taken accordingly." The area "between the blue lines" was in front of the 25th Infantry Division's main line of defense, an area about to be occupied by the enemy. Two things are clear: actions had been taken in conjunction with the Korean National Police to clear the civilians out of the danger area, and those actions were intended to ensure that noncombatants would not find themselves in harm's way when the advancing NKPA subsequently made contact along the Division's front. After the area was cleared, anyone caught in civilian clothes and suspected of being an enemy agent was to be turned over to the Counter-Intelligence Corps immediately and not to the Korean Police. There is nothing to suggest any summary measures were considered against refugees or people dressed like refugees. The 25th Infantry Division was not located in the vicinity of No Gun Ri during the last week of July 1950.

Finding. Based upon the available evidence, and despite some conflicting statements and misunderstandings, the U.S. Review Team concluded that U.S. commanders did not issue oral or written orders to shoot and kill Korean civilians during the last week of July 1950 in the vicinity of No Gun Ri.

A veteran stated that soldiers could have misunderstood the order not to let refugees pass or to stop refugees. Some veterans did believe that if a civilian would not stop, they could use deadly force to prevent civilians from passing.

Some veterans stated that there was an order to shoot civilians at No Gun Ri but had no information to support their assertions. These soldiers did not know who gave the order, did not hear the order, did not know when the order was given, and personally did not receive the order. As a result, the U.S. Review Team concluded that these veterans assumed that an order was given because artillery and mortars were fired. The U.S. Review Team also considered media statements quoting veterans who claimed that an order to shoot Korean civilians was given at No Gun Ri. The U.S. Review Team was unable to confirm these reports because the witnesses either were not at No Gun Ri at the time or refused to speak to the U.S. Army.

Although the U.S. Review Team found four references (entry in the 8th Cavalry Regiment Message Log, 25th Infantry Division Commander's order, Colonel Rogers' memorandum, and an extract from the U.S. Navy's Aircraft Carrier Valley Forge Activity Summary) discussing actions against civilians, it did not find evidence of an order given to soldiers by a U.S. commander, orally or in writing, to kill Korean civilians in the vicinity of No Gun Ri in the last week of July 1950.

 

H. Key Issue 8: Number of Korean Deaths and Injuries

Background. After taking the statements of U.S. veterans and securing the professional evaluation of the August 6, 1950, aerial reconnaissance photograph by the National Imagery and Mapping Agency and the Armed Forces Institute of Pathology, the U.S. Review Team asked the ROK Review Team to provide information on the number of casualties. The U.S. Review Team's research revealed no official records of refugee deaths or injuries in the vicinity of No Gun Ri between July 26 and July 29, 1950. The initial Associated Press articles reported hundreds of people killed. Korean witness estimates range between 60–100 dead in the double tunnel and 50–150 dead or injured from strafing / bombing. Several U.S. veterans describe a lower number of dead or injured civilians. The soldiers did not check the areas where civilians came under fire to determine whether there were dead bodies, and some estimates appear to be guesswork or to be based on recollections not related to No Gun Ri.

At three different meetings, ROK officials reported an unverified number of 248 casualties, which they stated was provided to them by the Yongdong County Office. But the ROK Review Team acknowledges that the estimated figure of 248 is not considered factual and will have to be substantiated by an additional investigation at some future date by the ROK government.

Finding. Based on the available evidence, the U.S. Review Team is unable to determine the number of Korean civilians who were killed or injured in the vicinity of No Gun Ri. During their investigation, the ROK Review Team reported that the Korean survivors' organization claimed an unverified number of 248 South Korean civilians killed, injured, or missing in the vicinity of No Gun Ri between July 25 and 29, 1950. This report was recorded by the Yongdong County Office. The ROK Steering Group, at a ROK-U.S. Steering Group meeting on December 6–7, 2000, in Seoul, ROK, reiterated the claim of 248 casualties.

The actual number of Korean casualties cannot be derived from the U.S. veteran statements and Korean witness statements. The U.S. Team believes that number to be lower than the Korean claim. An aerial reconnaissance photograph of the No Gun Ri area taken on August 6, 1950, shows no indication of human remains or mass graves in the vicinity of the No Gun Ri double railroad overpass. Korean burial customs, farming in the area, lack of reliable information, wartime disruptions of the countryside, and the passage of time preclude an accurate determination of the numbers involved.

Conclusion

During late July 1950, Korean civilians were caught between withdrawing U.S. forces and attacking enemy forces. As a result of U.S. actions during the Korean War in the last week of July 1950, Korean civilians were killed and injured in the vicinity of No Gun Ri. The U.S. Review Team did not find that the Korean deaths and injuries occurred exactly as described in the Korean account. To appraise these events, it is necessary to recall the circumstances of the period. U.S. forces on occupation duty in Japan, mostly without training for, or experience in, combat were suddenly ordered to join ROK forces in defending against a determined assault by well-armed and well-trained NKPA forces employing both conventional and guerilla warfare tactics. The U.S. troops had to give up position after position. In the week beginning July 25, 1950, the 1st Cavalry Division, withdrawing from Yongdong toward the Naktong River, passed through the vicinity of No Gun Ri. Earlier, roads and trails in South Korea had been choked with civilians fleeing south. Disguised NKPA soldiers had mingled with these refugees. U.S. and ROK commanders had published a policy designed to limit the threat from NKPA infiltrators, to protect U.S. forces from attacks from the rear, and to prevent civilians from interfering with the flow of supplies and troops. The ROK National Police were supposed to control and strictly limit the movements of innocent refugees.

In these circumstances, especially given the fact that many of the U.S. soldiers lacked combat-experienced officers and Non-commissioned officers, some soldiers may have fired out of fear in response to a perceived enemy threat without considering the possibility that they might be firing on Korean civilians.

Neither the documentary evidence nor the U.S. veterans’ statements reviewed by the U.S. Review Team support a hypothesis of deliberate killing of Korean civilians. What befell civilians in the vicinity of No Gun Ri in late July 1950 was a tragic and deeply regrettable accompaniment to a war forced upon unprepared U.S. and ROK forces.