Ducks turned out to be the problem.
In 2003 a team led by Arnold Schecter of the University of Texas School of Public Health sampled food animals around Bien Hoa City, about 35 kilometres from Ho Chi Minh City. They measured TCDD, the dioxin that contaminated Agent Orange, at up to 343 parts per trillion in duck tissue. A normal background reading sits below 0.1. Free-range chickens carried it. So did fish, and one unlucky toad.
None of those animals had been sprayed with anything. They foraged in the mud of a city where, in 1971, thousands of gallons of herbicide leaked out of storage tanks at an American airbase, and the mud held onto it for three decades.
What the three million figure covers
Red Cross estimates place roughly three million Vietnamese among those affected by dioxin exposure, including at least 150,000 children born with serious birth defects, and the Aspen Institute’s Agent Orange in Vietnam program treats those numbers as its working baseline. Some of those children belong to families with no direct exposure to the chemical at all, the harder thing to explain. Precision is not on offer regardless. Vietnamese medical records from the war years are patchy, and pinning any individual birth defect on a chemical sprayed sixty years ago sits beyond what the evidence can do.
Counting has been attempted anyway. In a 2023 report for the United States Institute of Peace, Susan Hammond and Đặng Quang Toàn cite figures from the Vietnam Association of Victims of Agent Orange: 150,000 second-generation victims and 35,000 in the third. Those are advocacy numbers, not epidemiology, and they are the best anyone has.
Why the chemical is still here
Agent Orange broke down fast. Its two herbicide components degraded within weeks, and the contaminant riding along with them did not.
Dioxin’s half-life in a human body runs 11 to 15 years and can stretch to 20, per figures compiled by the same Aspen programme. In sunlight, on leaf and soil surfaces, it degrades in one to three years. Buried below the surface, or settled into river sediment, its half-life passes a century. That last number explains why a chemical last sprayed in 1971 remains a live public health problem in 2026.
How exposure reaches people who were never near it
Food.
A 2010 risk assessment of two wards of Bien Hoa City, published in the International Journal of Environmental Research and Public Health, surveyed 400 households on what they ate and matched it against dioxin readings in local soil, mud, milk and blood. Residents eating locally raised freshwater fish, free-range chicken, duck and beef were taking in dioxin at daily rates far above the tolerable intake set by the World Health Organization. Bottom-feeding fish were among the worst offenders, which follows once you know the compound sinks into sediment and stays put.
That answers the question the headline number raises. A child born in 2010 to parents who never saw the war can still carry a body burden of a chemical the American military brought to Vietnam in the 1960s, because it went into the ground, the ground fed the ducks, and the ducks fed the family.
What the studies can and cannot show
That is one pathway. The older, more contested question is whether a parent’s own wartime exposure altered the children born afterward, and arguments over that link have run for fifty years. The most cited attempt to settle it is a systematic review and meta-analysis by Anh Ngo, Richard Taylor, Christine Roberts and Tuan Nguyen, which appeared in the International Journal of Epidemiology in 2006. It pooled 22 studies and found parental exposure associated with elevated risk that grew stronger as exposure intensity and duration rose.
One meta-analysis is one meta-analysis. Its authors flagged substantial heterogeneity between the studies they pooled, and a commentary printed alongside it in the same journal observed that the question had been contested since the 1960s without resolution. Separately, a ProPublica analysis of Department of Veterans Affairs data found the odds of a birth defect were more than a third higher among American veterans exposed to the defoliant, drawn from records the agency had collected for decades and never examined.
Cleanup, and the weeks it went dark
Contamination concentrated in a handful of former storage bases. Da Nang was finished in 2018. Bien Hoa, the largest remaining hot spot, holds roughly 500,000 cubic metres of contaminated soil and sediment, per a fact sheet from the US embassy in Hanoi, with a ten-year remediation programme costed at up to $450 million and funded jointly by USAID and the Pentagon.
In February 2025 it stopped. Secretary of State Marco Rubio’s freeze on foreign assistance caught the project, and ProPublica reported that crews abandoned the site leaving open pits of contaminated soil under tarpaulins, some of which blew off in the wind, weeks before Vietnam’s rainy season. The US ambassador in Hanoi sent a cable to Washington. Congress and USAID’s inspector general each received a whistleblower complaint.
Work resumed. As reported by the Associated Press, the State Department confirmed three dioxin remediation contracts were running again while declining to say how long the shutdown had lasted. Vietnam’s foreign ministry had already noted publicly that American war legacy funding amounts to some $200 million a year.
Ground broke on the largest single component in September 2025. Vietnamese state broadcaster VOV reported that the thermal treatment system heats contaminated soil to about 1,000°C, then burns off the resulting toxic gas at a destruction rate above 99 percent, with the whole site targeted for completion before 2030.
In March 2026, fourteen US senators wrote to the secretaries of state and defence urging continued American assistance on war legacy programmes in Vietnam. Soil is the easy part. It can be dug up, trucked, heated and handed back clean, and the Vietnam News Agency puts nearly half the base in that category already. There is no kiln for the third generation.
