Signs of impending communist action did not go unnoticed among the allied intelligence collection apparatus in Saigon. During the late summer and fall of 1967 both South Vietnamese and U.S. intelligence agencies collected clues that indicated a significant shift in communist strategic planning. By mid-December, mounting evidence convinced many in Washington and Saigon that something big was underway. During the last three months of the year intelligence agencies had observed signs of a major communist military buildup. In addition to captured documents (a copy of Resolution 13, for example, was captured by early October), observations of enemy logistical operations were also quite clear: in October the number of trucks observed heading south through Laos on the Hồ Chí Minh Trail jumped from the previous monthly average of 480 to 1,116. By November this total reached 3,823 and, in December 6,315. On 20 December Westmoreland cabled Washington that he expected the communists “to undertake an intensified countrywide effort, perhaps a maximum effort, over a relatively short period of time.”
Despite all the warning signs, however, the allies were still surprised by the scale and scope of the offensive. According to ARVN Colonel Hoang Ngoc Lung the answer lay with the allied intelligence methodology itself, which tended to estimate the enemy’s probable course of action based upon their capabilities, not their intentions. Since, in the allied estimation, the communists hardly had the capability to launch such an ambitious enterprise: “There was little possibility that the enemy could initiate a general offensive, regardless of his intentions.” The answer could also be partially explained by the lack of coordination and cooperation between competing intelligence branches, both South Vietnamese and American. The situation from the U.S. perspective was best summed up by an MACV intelligence analyst: “If we’d gotten the whole battle plan, it wouldn’t have been believed. It wouldn’t have been credible to us.”
From spring through the fall of 1967, the U.S. command in Saigon was perplexed by a series of actions initiated by the North Vietnamese and the Viet Cong in the border regions. On 24 April a U.S. Marine Corps patrol prematurely triggered a North Vietnamese offensive aimed at taking the airstrip and combat base at Khe Sanh, the western anchor of the Marine’s defensive positions in Quang Tri Province. By the time the action there had ended in May, 940 North Vietnamese troops and 155 Marines had been killed. For 49 days during early September and lasting into October, the North Vietnamese began shelling the U.S. Marine outpost of Con Thien, just south of the Demilitarized Zone or DMZ. The intense shelling (100–150 rounds per day) prompted Westmoreland to launch Operation Neutralize, an intense aerial bombardment campaign of 4,000 sorties into and just north of the demarcation line.
On 27 October, an ARVN battalion at Song Be, the capital of Phuoc Long Province, came under attack by an entire North Vietnamese regiment. Two days later, another North Vietnamese Regiment attacked a U.S. Special Forces border outpost at Loc Ninh, in Binh Long Province. This attack sparked a ten-day battle that drew in elements of the U.S. 1st Infantry Division and the ARVN 18th Division and left 800 North Vietnamese troops dead at its conclusion.
The most severe of what came to be known as “the Border Battles” erupted during October and November around Dak To, another border outpost in Kontum Province. The clashes there between the four regiments of the 1st North Vietnamese Division, the U.S. 4th Infantry Division, the U.S. 173rd Airborne Brigade, and ARVN infantry and Airborne elements, lasted for 22 days. By the time the fighting was over, between 1,200 and 1,600 North Vietnamese and 262 U.S. troops had lost their lives. MACV intelligence was confused by the possible motives of the North Vietnamese in prompting such large-scale actions in remote regions where U.S. firepower and aerial might could be applied indiscriminately. Tactically and strategically, these operations made no sense. What the communists had done was carry out the first stage of their plan: to fix the attention of the U.S. command on the borders and draw the bulk of U.S. forces away from the heavily populated coastal lowlands and cities.
Westmoreland was more concerned with the situation at Khe Sanh, where, on 21 January, a force estimated at between 20,000–40,000 North Vietnamese troops had besieged the U.S. Marine garrison. MACV was convinced that the communists planned to stage an attack and overrun the base as a prelude to an all-out effort to seize the two northernmost provinces of South Vietnam. To deter any such possibility, he deployed 250,000 men, including half of MACV’s U.S. maneuver battalions, to the I Corps Tactical Zone.
This course of events disturbed Lieutenant General Frederick Weyand, commander of U.S. forces in III Corps, which included the Capital Military District. Weyand, a former intelligence officer, was suspicious of the pattern of communist activities in his area of responsibility and notified Westmoreland of his concerns on 10 January. Westmoreland agreed with his estimate and ordered 15 U.S. battalions to redeploy from positions near the Cambodian border back to the outskirts of Saigon. When the offensive did begin, a total of 27 allied maneuver battalions defended the city and the surrounding area. This redeployment may have been one of the most critical tactical decisions of the war.