In depth | Operation Hong Kil Dong — why seven battalions lost the enemy main body

2026-09-23 · 1967-07-09 – 1967-08-26

Overview

Part 1: Operation Hong Kil Dong (July–August 1967) traced the big picture: a converging attack and a trail that went cold. This piece takes the decisions apart. Seven Korean battalions closed on the 'Hub', a mountain base area north-west of Tuy Hoa, trapped several enemy companies, and never caught the main body of the North Vietnamese 95th Regiment. The US Army's own study holds the operation up as a model Korean result, yet in the same chapter criticises Korean reaction speed and helicopter planning [1]. Below, weather and moon data reread the ground, and the scenario's phases p1–p7 (plus hypothetical injects x1–x3) show where "the enemy we trapped" and "the enemy we lost" parted ways.

Aerial photograph of the Tuy Hoa Army airfield at Phu Hiep in July 1967: a runway, taxiway and parking apron of steel matting laid over flat coastal sand — Photo 1. Tuy Hoa Army airfield, Phu Hiep (15 July 1967). In May–June the US 577th Engineer Battalion laid M8A1 matting on the runway and taxiway; it was part of the coastal base that sustained helicopter resupply and air assaults during the operation — Office of History, US Army Corps of Engineers, public domain (US federal government work), via Wikimedia Commons [11]

Strategic and operational context

In 1967 South Korea had about 48,000 troops in Vietnam [1] (over 300,000 served in all [15]). The Capital Division held the Qui Nhon area; the 9th Division, at Ninh Hoa since late 1966 and counting Hong Kil Dong among its main operations [14], kept its 28th Regiment at Tuy Hoa on Highway 1 [1]. Guarding the populated coast was the basic mission. Oh Jak Kyo in March–April had linked the two divisions along the highway; Hong Kil Dong went into the mountains behind it [4].

The 95th Regiment came south in 1964 with about 2,000 men and became Phu Yen's main regular unit. Fighting with the 101st Airborne cut it to about 900 by late 1966, and in early 1967 it withdrew into the hills, working with two VC battalions [2]. In November 1966 Operation Geronimo I had trapped part of it in the Ky Lo Valley and west of the Hub without driving it out of the province [3][5].

Lt. Gen. Chae Myung-shin commanded this corps-level operation himself [4]. Two intents pulled against each other: break the regulars in the mountains, and do not strip the coastal plain. The size of the force and the end date were both set between them. In the background was the 3 September presidential election, which the 95th came back to the coast in August to disrupt [2].

Terrain and weather (OAKOC)

The battle area is the high ground inside the Tuy Hoa–Dong Tre–Cung Son triangle [17], with the Hub as its central massif. To its north-west the Ky Lo Valley climbs to the Phu Yen–Phu Bon border; one historian describes it as a large enemy base and staging area with forges and a hospital-dispensary [3].

Figure 1. Detailed map (approximate). Terrain from the training scenario map; axes and dates from sources 1, 2 and Part 1; battalion positions are estimated — own work

Observation and fields of fire: Jungle defeated air observation and cut fields of fire to a few dozen metres: the case where FM 3-90 prescribes search and attack against an enemy targetable only "by … physical search" [8].
Avenues of approach: Ridges and valley floors, one route south-east from Dong Tre and one north-east from Cung Son. The same valleys were the enemy's exits.
Key terrain: The Hub, and the eastern ridges overlooking the coastal plain.
Obstacles: Steep slopes, jungle, streams; one to two kilometres a day was close to what the ground allowed. Landing zones were small and few.
Cover and concealment: Base camps and caves with unknown numbers of entrances, the problem at p2.

Weather and light. No daily observations were found. ERA5 reanalysis near the Hub (13.25°N, 109.05°E) gives highs of 29.6–34.5°C (mean 32.4°C) and mean lows of 25.2°C over the 49 days, with about 128 mm of rain; only 24 August (25 mm) exceeded 15 mm [9]. The heavy rains come after September, so this was hot, fairly dry campaigning in which water weighed as much as ammunition. A reanalysis grid misses mountain microclimates; read these as trends. Daylight ran from 05:22 to 18:14 on 9 July (UTC+7) [9]. Full moons fell on 21 July and 20 August, the new moon on 6 August [10]: the loss of contact in early August coincides with the darkest nights, though when the enemy actually left is unrecorded.

Organisation and command relationships

The sources do not give battalion numbers; regimental assignments are estimates from garrison locations and other records.

Role	Unit	Size and task	Basis
Command	ROK Forces Vietnam (Lt. Gen. Chae Myung-shin)	Directed the corps-level operation	[4]
Northern attack	Capital (Tiger) Division	4 bns, attacking south-east from south of Dong Tre; regiment unconfirmed (26th Regiment elements estimated)	[Part 1][2], est.
Southern attack	9th (White Horse) Division	3 bns, attacking north-east from Cung Son; Tuy Hoa-based 28th Regiment estimated	[Part 1][1], est.
Eastern block	ARVN 22nd Division, 47th Regiment	2 bns blocking at the edge of the coastal plain	[Part 1][2]
Fires	USAF B-52s, ROK divisional artillery	Pre-assault strike on 9 July; fires on suspected sites thereafter	[1][Part 1]
Aviation	US 10th Combat Aviation Battalion	Direct support 0700 9 July – 2400 31 August; veterans name the 129th Assault Helicopter Co, 180th Assault Support Helicopter Co (CH-47) and 403rd Transportation Det	[17][6]
Enemy command	NVA 5th Division HQ, rear services	Assessed to be in the Ky Lo Valley	[Part 1]
Enemy main body	95th Regiment (4th, 5th, 6th Bns)	About 900 at end-1966, reinforced in 1967; total in the western mountains assessed at about 1,800	[2][17]
Enemy partners	85th LF Bn, 30th MF Bn	VC units working with the 95th	[2]
Enemy south	188th Regiment	Assessed on the Phu Yen–Khanh Hoa border	[Part 1]

Nine battalions could never seal the whole Hub. The Korean strength was the dense cordon, with "less than 10 meters" between men, as a US commander noted [1]; that works around a hamlet, not a massif tens of kilometres around. And the gap between the estimate of about 1,800 enemy and the few companies met is the central question.

Sequence of events

9 July — strike and start

B-52s opened the operation, followed that morning by a "multi Regimental assault" [17] from south of Dong Tre and from Cung Son. First contacts were light [Part 1].

10 July to mid-July — the slow ring

Both divisions combed ridges and valleys at one to two kilometres a day or less [Part 1]; contemporary film shows heliborne lifts, small-unit searches and casualty evacuation [13][6]. Enemy companies went to ground in camps and caves and were sealed in and reduced. The US Army study, from Korean reports, gives 408 enemy losses for 9–31 July [1].

Late July — through the Hub

All seven battalions swept east through the Hub. Resistance was strong, then almost vanished [Part 1].

Early August — contact lost

For about two weeks there was almost no contact. Phase 1 closed with the clearing of bypassed pockets, date unrecorded [Part 1]. The CMH account dates a raid on Hieu Xuong district headquarters, beside Tuy Hoa Air Base, to 9 August, though the date sits awkwardly in its narrative [Part 1].

12–25 August — phase 2 search

The Capital Division climbed the Ky Lo Valley to the Phu Bon border by the morning of the 25th without finding the division headquarters; the 9th found no trace of the 188th in the south [Part 1]. The period's heaviest rain, about 36 mm, fell on 24–25 August [9].

26–31 August — the end

The battalions returned to the coast; the end date varies by source, and US aviation support ended at 2400 on 31 August [17]. The reinforced 95th then came back to the Tuy Hoa coast and suffered several hundred losses against the Capital Division's 26th Regiment [2]. That fighting around coastal hamlets, which caused heavy harm to residents, is outside this article and the scenario.

Animated map. Six phases appear in turn: the 9 July strike area and two converging axes; three enemy camps ringed inside the Hub in mid-July; the late-July sweep east against a rearguard; the estimated south-east move of enemy remnants and the district headquarters raid in early August; the 12–25 August searches up the Ky Lo Valley and along the southern border; and in late August the return to coastal bases while the 95th reappears near the coast. The still frame shows the final situation — Figure 2. Manoeuvre by phase (18-second loop; with reduced motion only the final situation is shown). Approximate — own work, based on sources 1, 2 and Part 1

Decision points in depth

The phases take the view of the Capital Division's four-battalion headquarters. Each "actual decision" is inferred from outcomes; no record of the commanders' reasoning was found.

Decision flow chart. Top row: p1 How fast to push, p2 Enemy in the caves, p3 Through the Hub, p4 Empty mountains; bottom row: p5 Trail goes cold, p6 The Ky Lo Valley, p7 At the border, each with the actual decision. Dashed boxes are hypothetical injects x1 Gap in the ring, x2 Western foothills, x3 Stretched supply. Below, a bar chart of daily rainfall from 9 July to 26 August totals about 128 mm, with the peak of 25 mm on 24 August; full moons on 21 July and 20 August and the new moon on 6 August are marked — Figure 3. Decision flow, rainfall and moon phase. Phases from the training scenario, rainfall from ERA5 reanalysis [9], moon from NASA [10] — own work

p1 How fast to push (9 July)

Known: Bombing results and regiment's location unknown; contacts light.
Options: (a) comb ridges and valleys at 1–2 km a day; (b) heli-lift a battalion into the Hub's centre; (c) wait for patrols to assess the bombing; (d) race along the ridgelines to close the ring early.
Decision and result: (a), the Korean norm: cordons "very slow and deliberate, not a rock is left unturned", in a US general's words [1]. What was inside was caught; the main body had time to leave.
Counterfactual: (d) closes sooner but opens intervals between units moving at different speeds [8]; (b) needed a location, and US evaluators judged Korean airmobile planning poor [1]. Inject x1 (gap in the ring) tests the cost of slowness.

p2 The enemy in the caves (mid-July)

Known: A company in caves, entrances unknown; a timetable to keep.
Options: (a) trap, reduce and clear; (b) bypass for follow-on forces; (c) assault at once; (d) air strike and move on.
Decision and result: Close to (a) [Part 1], some pockets bypassed and cleared later. Slow but sparing of lives.
Counterfactual: (d), reduction by fires alone, is FM 3-90's least preferred method, costly and often incomplete [8]; (c) risked heavy losses blind.

p3 Through the Hub (late July)

Known: Camps mostly cleared; a dug-in rearguard on the eastern ridges; the 9th alongside, ARVN in the east.
Options: (a) all seven battalions sweep east; (b) 9th Division blocks, Capital Division attacks; (c) fix with one battalion, envelop over the northern ridges with two; (d) halt and fire first.
Decision and result: (a). Resistance soon faded [Part 1]; a rearguard that holds briefly and vanishes may be screening a main body already moving.
Counterfactual: (c) might have cut the rearguard off, at the risk of crossed coordination lines between converging forces [8]; (b) would have closed southern exits.

p4 Empty mountains (early August)

Known: Two weeks without contact; tired troops; pockets behind; the 95th missing.
Options: (a) clear the pockets and close phase 1; (b) push east to the plain; (c) spread patrols wide to regain contact; (d) rest battalions in rotation.
Decision and result: (a) [Part 1], consistent with the rule that bypassed enemy must be kept under contact until handed off [8], but it put the known enemy ahead of the lost one.
Counterfactual: (c) is the key alternative: the more uncertainty, the larger the reconnaissance force [8]. A wide patrol net might have caught the 95th moving coastward, though that is hindsight.

p5 The trail goes cold (12 August)

Known: Two weeks of silence; a raid reported on the coast; headquarters placed the 5th Division command post in Ky Lo and the 188th on the southern border.
Options: (a) four battalions into Ky Lo; (b) main body in the Hub, patrols west of Tuy Hoa; (c) two battalions down to block the approaches to the coast; (d) stay and fire on suspected sites.
Decision and result: (a) plus a southern border search; both empty [Part 1]. Ky Lo had held an enemy base in 1966 [3], so the logic was sound; the weak point was stale intelligence, since the coastal raid pointed the other way.
Counterfactual: (c) fits the intent of protecting the coast; with the 95th actually heading there [2], the most "defensive" option might have met the most enemy. Inject x2 turns on this.

p6 The Ky Lo Valley (mid-August)

Known: Nothing found; landing zones shrinking, resupply gaps growing.
Options: (a) keep combing to the border; (b) turn back to the Hub; (c) leapfrog companies by helicopter; (d) halt and search the valley floor.
Decision and result: (a); the border was reached on the 25th [Part 1], making the negative result certain.
Counterfactual: (c) is textbook search and attack [8], but narrow LZs and strained helicopter support [1] made it hard. Inject x3 (stretched supply) probes that limit.

p7 At the border (25 August)

Known: No headquarters, no main body; troops exhausted; coastal units anxious.
Options: (a) end it, all battalions home; (b) leave two battalions west of Tuy Hoa; (c) sweep the Hub again on the way back; (d) withdraw but leave long-range patrols and air observation.
Decision and result: (a). Within days the 95th was back near the coast [2].
Counterfactual: (d) was the cheapest early warning, as doctrine advises when the finishing force cannot be committed [8], and it meets the US criticism that Korean units idled too long after operations [1].

Doctrine revisited

Search and attack (FM 3-90, paras 4-50 to 4-63). Doctrine splits the force into reconnaissance, fixing and finishing elements [8]. In phase 1 all seven battalions did everything at once, so nothing mobile chased what got out. Para 4-62 wants a finishing force able to strike "before they can break contact"; the slow tempo ran the other way.

Encirclement (appendix C). Encirclements take large forces and time (C-10); intervals open between units moving at different rates and the enemy looks for them (C-15); mobile forces may hold only key terrain if they can detect breakouts (C-17) [8]. Nine battalions could never form a continuous line around a massif, so C-17 was the realistic model, with posts, ambushes and planned fires on every exit.

Reduction by fires (C-19, C-20). A Korean hill attack reportedly took about a month to prepare with B-52s, artillery and tanks [1]: lives saved, time spent, results hard to confirm.

Bypass and contact (para 12-27). Bypassed enemy stays under contact until handed off [8]. p2 and p4 honoured this; regaining contact with the main body seems never to have been anyone's task.

The enemy's method. As after Geronimo I, the 95th "went to ground" during the sweep and returned when the pressure lifted [3][2].

Source criticism

Issue	Source A	Source B	Assessment
End date	26 Aug (Korean accounts, Part 1)	28 Aug (US Army study [1]); 31 Aug (end of US aviation support [17])	Ground operation and aviation support ended separately; 26 vs 28 likely reflects when units finished returning. Korean originals not seen
Enemy losses claimed, early	500+ by 25 July (Korean announcement, via Part 1)	408 for 9–31 July (US Army study [1])	The later-dated B is lower; A was probably a provisional figure at the time of release
Enemy losses claimed, final	700+ and 450+ weapons (Chae's memoir, via Part 1)	638, with 359 individual and 98 crew-served weapons (US Army study [1])	B's weapons total, 457, fits A's "450+". B has the clearer counting basis. Both derive from Korean reports and are unverified claims
Enemy strength	About 1,800 in the western mountains (aviation report summary [17])	95th about 900 at end-1966 (US Army historian [2])	A is an estimate for several units, B for one regiment; no contradiction
Assessment of Korean forces	Westmoreland praised morale and professional competence [1]	Slow reaction, poor helicopter planning, long inactivity after operations [1]	Both in the same chapter: praise looks at outcomes (exchange ratios), criticism at process (tempo and manoeuvre)

The US study's ratios (15 to 1, 24 to 1) were the allied yardstick of the day, one an analysis of Geronimo I finds showed no lasting gain [3]; they appear here only to show where sources diverge. MacGarrigle's CMH volume [16] and the aviation battalion report [17] could not be opened (403); reading them would settle many battalion numbers and dates.

Aerial photograph of Tuy Hoa Air Base in November 1967: a broad steel-matted apron, runway and maintenance facilities spread across coastal dunes — Photo 2. Tuy Hoa Air Base (28 November 1967), just after the 577th Engineer Battalion finished matting the maintenance apron. Hieu Xuong district headquarters lay beside this base — unknown photographer (NARA 111-CCV-34-CC80857), public domain (US federal government work), via Wikimedia Commons [12]

Lessons

Plan exit surveillance with the cordon: watch and pre-plan fires on every valley exit (FM 3-90, C-15).
Separate clearing the known enemy from finding the lost one.
Weigh the age of intelligence: check the newest report first.
Return to the higher intent: when the enemy cannot be found, protect what matters; here, the coast.
Plan the ending: leave small surveillance elements behind.
Supply sets the search radius: build landing-zone security into the plan.

Open questions

Battalion numbers and regiments (memorial records hinting at 9th Division 30th Regiment involvement could not be opened).
When and by which route the 95th's main body left the ring.
Whether 9 August is the right date for the Hieu Xuong raid.
The Korean divisions' original combat reports, cited in one study's bibliography [4] but not publicly available.

Further viewing

ROK Operation Hong Kil Dong 1967

Video: field footage of Operation Hong Kil Dong, 1967 (about 14 minutes), shot by a US Army combat cameraman, with comments from members of the US helicopter units that supported the operation · 221st Signal Company (US Army pictorial company veterans), 2021 [6]

Daehan News No. 654, news from Vietnam

Video: Daehan News No. 654, news from Vietnam (produced 22 December 1967, about 2 minutes), showing Capital Division operations and village reconstruction later that year · KTV Archive (Korea Policy Broadcasting Service), 2016 [7]

Footage: Operations of the Tiger Division in Vietnam (probably 1967, about 3 minutes; date and place estimated from equipment) — War Memorial of Korea Open Archives, KOGL Type 1, via Wikimedia Commons [13]

Sources

Stanley R. Larsen and James L. Collins, Jr., Allied Participation in Vietnam, Vietnam Studies, Department of the Army, 1975 (CMH Pub 90-5), chapter VI, "The Republic of Korea" (Göttingen University library mirror of CMH). https://webdoc.sub.gwdg.de/ebook/p/2005/CMH_2/www.army.mil/cmh-pg/books/vietnam/allied/ch06.htm (accessed 2026-10-14)
Erik Villard, "The NVA's Plan for a Great Escape", Vietnam, February 2018 (HistoryNet, 2 January 2018). https://www.historynet.com/nvas-plan-great-escape/ (accessed 2026-10-14)
Robert J. Thompson, "Geronimo: Strategy Lessons from History's Periphery", The Strategy Bridge, 4 December 2016. https://thestrategybridge.org/the-bridge/2016/12/4/geronimo-strategy-lessons-from-historys-periphery (accessed 2026-10-14)
Heo Dong-uk, "Reward and Lesson of Ojakgyo and Honggildong Military Operations in the Vietnam War", Journal of Korean Veterans Studies 19(2), 2020, pp. 9–42 (KCI abstract). https://www.kci.go.kr/kciportal/landing/article.kci?arti_id=ART002604120 (accessed 2026-10-14)
2nd Brigade, 101st Airborne Division association, Weekly History 06–12 November (2016), citing HQ 2-502 Infantry After Action Report, Operation Geronimo I (10 December 1966). https://2ndbde.org/weekly_history/2016/06_november-12_november_2016.pdf (accessed 2026-10-14)
221st Signal Company (US Army pictorial company veterans), "ROK Operation Hong Kil Dong 1967" (YouTube), 2021. https://www.youtube.com/watch?v=i6EnC92PfUE (accessed 2026-10-14)
KTV Archive (Korea Policy Broadcasting Service), "Daehan News No. 654 – news from Vietnam" (YouTube), 2016 (originally produced 22 December 1967). https://www.youtube.com/watch?v=c62PwDi9YSg (accessed 2026-10-14)
Department of the Army, FM 3-90 Tactics, 1 May 2023 (chapter 4 paras 4-50 to 4-63; para 12-27; appendix C paras C-9 to C-20 and C-26). https://armypubs.army.mil/epubs/DR_pubs/DR_a/ARN38160-FM_3-90-000-WEB-1.pdf (accessed 2026-10-14)
Open-Meteo, Historical Weather API (ERA5 reanalysis), grid 13.25°N 109.05°E, daily temperature, rainfall, sunrise and sunset, 9 July – 26 August 1967. https://archive-api.open-meteo.com/v1/archive?latitude=13.25&longitude=109.05&start_date=1967-07-09&end_date=1967-08-26&daily=temperature_2m_max,temperature_2m_min,precipitation_sum,sunrise,sunset&timezone=Asia%2FBangkok (accessed 2026-10-14)
NASA Goddard Space Flight Center, Phases of the Moon: 1901 to 2000 (July–August 1967). https://eclipse.gsfc.nasa.gov/phase/phases1901.html (accessed 2026-10-14)
Wikimedia Commons, File:Phu Hiep, July 1967.png (USACE Office of History, public domain). https://commons.wikimedia.org/wiki/File:Phu_Hiep,_July_1967.png (accessed 2026-10-14)
Wikimedia Commons, File:Tuy Hoa air base in November 1967.jpg (NARA 111-CCV-34-CC80857, public domain). https://commons.wikimedia.org/wiki/File:Tuy_Hoa_air_base_in_November_1967.jpg (accessed 2026-10-14)
Wikimedia Commons, File:Operation of the Fierce Tiger Division of the Republic of Korea Armed Forces in Vietnam, 1967.webm (War Memorial of Korea Open Archives, KOGL Type 1). https://commons.wikimedia.org/wiki/File:Operation_of_the_Fierce_Tiger_Division_of_the_Republic_of_Korea_Armed_Forces_in_Vietnam,_1967.webm (accessed 2026-10-14)
Korea Democracy Foundation Open Archives, "White Horse Division in Vietnam: operations and daily life" (photo collection note, in Korean). https://archives.kdemo.or.kr/photo-archives/view/00755902 (accessed 2026-10-14)
National Institute of Korean History, Uri History Net, "The Vietnam War 1954–1975" (in Korean). https://contents.history.go.kr/mobile/kc/view.do?levelId=kc_i502200&code=kc_age_50 (accessed 2026-10-14)
George L. MacGarrigle, Combat Operations: Taking the Offensive, October 1966 to October 1967, US Army Center of Military History, 1998, pp. 311–314. https://history.army.mil/portals/143/Images/Publications/catalog/91-4.pdf (access attempted 2026-10-14; 403, text not seen — Part 1's cross-check followed)
10th Combat Aviation Battalion, Combat After Action Report, Operation Hong Kil Dong, 1967, DTIC AD0387534. https://apps.dtic.mil/sti/pdfs/AD0387534.pdf (access attempted 2026-10-14; 403, search summary only)