Substantial Transformation and Vietnam Origin: What the Rule Actually Says

The sentence everyone cites lives in 19 C.F.R. section 134.1(b), and its operative content is short: further work or material added to an article in another country must effect a substantial transformation before that other country can be treated as the country of origin. It sets no threshold, names no percentage, and lists no qualifying operations.

There is a limit written into that same definition that almost never survives the retelling. It applies within the meaning of that part, and Part 134 is the marking rules — the rules about what the label has to say. Where duty is being assessed, the substantial-transformation standard CBP applies comes out of case law rather than out of the marking definition, and goods claiming USMCA treatment are determined under a different set of rules in Part 102 entirely.

For a buyer the practical reading is that quoting the regulation is not the same as having an answer. The regulation tells you the test exists. Which facts satisfy it for your product is a question about your product, and the material that answers it is the production record your supplier can or cannot produce.

Vietnam's non-preferential origin regime runs on the same principle rather than on a number: origin is the country where the last processing operation that substantially changed the goods took place. The criteria that decide it sit in a product-specific rules list organised by HS code, in Annex I of Circular 05/2018/TT-BCT, where a heading is satisfied either by a change in tariff classification — at the 2, 4 or 6-digit level — or by a local value-content percentage. Where a percentage exists at all, it belongs to that heading, and the statutory text itself carries no general figure.

The base instrument is still Decree 31/2018/NĐ-CP. A drafting group for a replacement was set up in 2025 and a second draft went to consultation on 2025-05-23, but according to the Ministry of Industry and Trade no replacement decree has been issued — so anyone working from “the new Vietnamese origin decree” is working from a draft.

Article 9 of that decree lists the operations to be disregarded when origin is determined. It is worth reading as a list of things a factory can do without conferring origin:

If a supplier's list of local steps reads like the list above, the origin claim has a problem regardless of what the invoice says. The de minimis tolerance in Article 11 of the same decree does not rescue it either: non-originating materials up to 15% of the FOB value, or 15% by weight for textiles, are an allowance for materials that fail a classification-change rule. It is a tolerance inside a rule that has already been chosen, not a threshold that confers origin on its own.

This is the most common way a well-meaning compliance answer goes wrong. Someone finds a percentage in the origin rules, and a percentage is what they were looking for, so it becomes the answer to a question it was never written to answer.

The formulation that 30% local value added plus a last operation that is not simple processing makes goods Vietnamese circulates widely enough that suppliers repeat it back to buyers as settled law. It comes from a Made in Vietnam circular drafted in 2019 that was never issued, and according to Vietnamese official reporting the ministry was still submitting a draft decree on the same subject in mid-2026, on a voluntary self-declaration basis with no published percentage.

Two things follow. A supplier who cites 30% is citing a draft, which tells you where its compliance information comes from. And a domestic labelling rule, if one ever issues, would not be the rule that decides whether US customs accepts a Vietnamese origin claim — the two questions run on different instruments in different countries, and conflating them is how a factory ends up confidently wrong.

The test is applied to facts about production, so the facts that move it are physical. Where the inputs came from, and under which HS codes they entered. What the factory did to them, described as operations rather than as a job description. Whether the output falls under a different heading from the inputs. Whether local content, where a percentage rule applies to that heading, is documented rather than asserted.

This is why kit assembly is the recurring failure. Parts arrive under one classification, are screwed together, and leave under a heading that is often the same or adjacent, with local steps that map onto the disregarded list almost item for item. The paperwork can be immaculate and the claim still fails, because the claim was never about paperwork.

The same structural gap shows up in other Vietnamese rules that read simply on paper and turn out to be operational: data protection is easy to quote and hard to implement in the one screen where it actually lands. Origin has the same shape. The rule fits in a sentence; the compliance lives in records that have to exist before anybody asks for them.

Ask three questions in this order. The answer usually arrives before the third one.