How to Read a Tariff Classification Code, and Why One Digit Changes Your Duty

A Mexican tariff classification is eight digits long, and most operations treat all eight as a single field the customs broker fills in. They are not one field. They are a sequence, built in layers, and each layer is decided by a different rule and carries a different consequence. Change one pair of digits and you can change the duty rate, whether a NOM applies, whether the goods qualify for a trade preference, and, on the U.S. side, which Section 232 tier hits the full value of the product. The code is small. The exposure inside it is not.

This is a guide to reading that code the way the people who build it read it: layer by layer, knowing which rule governs each part and where the money actually moves. The goal is not to turn you into a classifier. It is to make you literate enough to recognize when a code looks wrong before it becomes a duty you never budgeted for.

How the code is built, layer by layer

Take a code like 6109.10.01.00 and read it from the left, because each step narrows the universe of goods.

The first two digits are the Capítulo (chapter). They place the product in one of the broad families of the Tarifa de la Ley de los Impuestos Generales de Importación y de Exportación (TIGIE). The first four digits are the Partida (heading). The first six are the Subpartida (subheading), and this six-digit level is the part the whole world shares: it comes directly from the Sistema Armonizado, the Harmonized System maintained by the World Customs Organization, which Mexico adopted through the Ley de los Impuestos Generales de Importación y de Exportación (LIGIE), published in the Diario Oficial de la Federación on June 7, 2022, implementing the System's Seventh Amendment.

Digits seven and eight are Mexico's. They complete the fracción arancelaria, the eight-digit national code that actually carries the arancel (the duty rate) and the non-tariff regulations and restrictions. This is the level where Mexican law makes its own distinctions beyond what the international system requires.

Then there are two more digits, the ninth and tenth: the Número de Identificación Comercial (NICO). The NICO is a fifth pair added after the fracción, created under Regla Complementaria 10a of the LIGIE, with its methodology published in the DOF on June 27, 2022. It exists for statistical and commercial-intelligence precision, and, importantly, it does not by itself set duties or non-tariff regulations. New NICOs take effect on fixed dates, January 1 and July 1 each year.

So the full ten-digit identifier is three things at once: a slice of an international standard (the first six digits), a national tariff decision (digits seven and eight), and a statistical tag (digits nine and ten). Reading it well means knowing which part you are looking at.

Why one digit moves money

Here is the part that matters for cost. The duty rate, the preference eligibility, and the regulatory burden do not attach to the whole code uniformly. They attach at the fracción level, the eight-digit code. Move the seventh or eighth digit and you can land on a fracción with a different IGI rate, a different NOM requirement, a different permiso previo, or a different padrón sectorial obligation. The product did not change. The legal treatment did.

This is not theoretical drift. Tariff rates on specific fractions change: in 2026 alone, the rates on 185 fractions of the TIGIE were modified. A code that carried one rate last year can carry another this year, which is precisely why classification cannot be a one-time setup filed and forgotten.

The ninth and tenth digits, the NICO, will not change your duty. But do not treat them as harmless. An inexact classification declaration, NICO included, is still an inexact declaration, and an inexact declaration is the kind of error that can suspend an operation in the Sistema Electrónico Aduanero until it is corrected. Right code, wrong tail, still a problem.

Where the Mexican and U.S. codes meet, and where they split

If you move goods across the border, you are reading two schedules, not one, and knowing where they agree and where they diverge prevents a costly assumption.

They agree at the top. Because both Mexico and the United States build on the same Harmonized System, the first six digits of a product's classification should align in both countries. A product is the same product; the world agreed on how to name it to six digits.

They diverge below that. Mexico extends the six digits to an eight-digit fracción plus a two-digit NICO, governed by the LIGIE. The United States extends the same six digits into the ten-digit Harmonized Tariff Schedule (HTSUS), administered by the U.S. International Trade Commission, with its own national breakouts and statistical suffixes. The duty rates, the preference rules, and the trade-remedy treatments are set independently on each side.

The practical warning: do not assume that because you have the Mexican fracción, you have the U.S. classification, or the reverse. The six-digit core carries over. Everything below it, including what you owe, is decided country by country.

The rules that decide the code, in plain language

Classification is not a guess or a search-and-pick exercise. It follows an ordered, binding logic, set out in Article 2 of the LIGIE.

First come the six Reglas Generales, the international General Rules of Interpretation. They are applied in strict order, each only when the one before it does not resolve the question, and they take you down to the six-digit subheading. They are the same rules every Harmonized System country applies, which is why the six-digit level is shared.

Then come the ten Reglas Complementarias, Mexico's national rules, which carry the classification from the six-digit subheading down to the eight-digit fracción and govern how the NICO is applied. The order matters as much as the content: classification is a sequence of exclusions, not a menu. A product belongs where the rules place it, not where it would be convenient to place it.

You do not need to memorize the rules to benefit from knowing they exist. The point is that a defensible classification can always be explained by reference to them. If no one in your operation can say why a product sits on the fracción it sits on, that is the gap worth closing.

Common misclassification patterns, and what each costs

Misclassifications cluster, and each pattern has a characteristic cost.

The most common is classifying by what a product is called rather than what it is. Commercial names and marketing categories are not classification criteria; composition, function, and degree of processing are. A code chosen from the product's name rather than its essential character is the classic source of a wrong fracción.

A second pattern is inheriting a classification. A code gets copied from a prior shipment, a supplier's document, or a similar-looking product, and never re-tested against the actual good. When the rate on that fracción later changes, or the product specification quietly shifts, the inherited code is wrong and no one notices until an audit does.

A third is treating the NICO as an afterthought. The fracción gets attention and the NICO gets defaulted, producing an inexact declaration even when the duty-bearing digits are right.

The cost of each runs along the same path: underpaid or overpaid duty, an omitted non-tariff regulation, exposure to assessments and penalties for inexact data, and, increasingly, a misaligned trade-remedy outcome. On the U.S. side, classification now determines which Section 232 tier applies, and since the April 2026 restructuring of that regime, the duty is calculated on the full customs value of covered articles, not just their metal content. A classification slip can move a product between tiers, and the difference is no longer rounding.

When to escalate

Some classifications are clear. Others are genuinely contestable, and knowing which is which is part of running the function well.

Treat a classification as routine when the product is well defined, the rules place it cleanly, and the code has been tested against the actual good rather than inherited. Escalate when the product could defensibly sit on more than one fracción, when the duty or preference consequence of the choice is material, when a supplier and your own team disagree, or when a rate or rule change has just landed on the heading in question. These are the cases where a formal review, and where appropriate a binding classification ruling, protects the operation.

This is also where a verification layer earns its place. A classification process that scores its own confidence, flagging the codes that are solid and surfacing the ones that are contestable, turns classification from a silent assumption into a managed risk. The codes you are sure of move. The codes you are not sure of get a second look before they become a liability.

A tariff classification is not a label. It is a chain of decisions, each governed by a rule, each carrying a consequence, and you have been signing operations against it. Learn to read it layer by layer, and the digit that would have cost you becomes the digit you caught.

If you want a classification review on the fractions that carry your largest duty exposure, talk to a Joffroy expert.

TRADE. UNDER CONTROL.