Ask someone new to classification how they'd find a product's HTS code, and the answer is usually some version of "search until I find one that matches." That's backwards. The schedule isn't a list you search — it's a hierarchy you narrow through, one level at a time, and the rules for narrowing are written down.
17,000+ Codes. One Is Correct.
The schedule gets bigger the deeper you go — your path through it only gets narrower.
Chapter
99 chapters
2 digits — the broad product category (Ch. 61 for knit apparel, Ch. 84 for machinery)
Heading
~1,244 headings
4 digits — narrows to a specific type of good within that chapter
Subheading
~5,000+ subheadings
6 digits — where the internationally harmonized part of the code stops
US Rate Line + Stat. Suffix
17,000+ codes
8–10 digits — the legal US duty rate and Census reporting number live here
#HTSClassification #CustomsBroker #TariffSchedule
The Part That Surprises People
The total number of possible codes gets bigger the more specific you go — 99 broad chapters fan out into roughly 1,244 headings, which fan out into more than 5,000 subheadings, which fan out into over 17,000 US rate lines and statistical suffixes. That's the opposite of what "narrowing down" sounds like.
For any one product, you're never actually choosing from all 17,000. You're walking exactly one path — one chapter, one heading, one subheading, one final code — and every step eliminates everything outside that branch.
Narrowing Isn't a Judgment Call — It's a Rule
The General Rules of Interpretation (GRI), printed at the front of the tariff schedule itself, govern how that narrowing happens. They're not optional guidance — they're the legal method:
- GRI 1 does most of the work: classification is determined by the heading text itself, plus any relevant section or chapter notes — read literally, in order
- GRI 2 extends a heading to cover an unfinished or unassembled version of the article it describes
- GRI 3 only applies when GRI 1 leaves goods classifiable under two or more headings — resolved by which heading is most specific, then by "essential character," then by whichever heading comes last in the schedule
- GRI 6 applies the same logic one level down, at the subheading
Most classifications resolve at GRI 1 — the heading text and the chapter notes settle it without needing to go further. GRI 3's essential-character test is the one that actually requires judgment, and it's the one most likely to produce a defensible-but-arguable answer worth documenting.
Where the Legal Weight Actually Sits
The first six digits — chapter, heading, subheading — are the internationally harmonized part of the code, set by the World Customs Organization and shared across more than 170 countries. The last four digits are US-specific: the 7th and 8th digits are the legal rate line that sets the duty owed, set by the USITC; the 9th and 10th are a Census Bureau statistical suffix that carries no duty consequence at all.
Two importers can agree on a product's 6-digit subheading and still land on different 8-digit rate lines — and only one of them is correct. The international harmonization ends well before the duty rate does.
How Declaro Reads This
Declaro's classification engine is trained on 220,000+ CBP CROSS rulings — the record of how CBP itself has actually applied GRI 1 through 6 to real products, at real subheadings, across every chapter. Instead of manually walking chapter notes and prior rulings to find the applicable branch, the engine surfaces the precedent that already resolved the same GRI question for a comparable product — so the elimination process that should take an afternoon doesn't.
Declaro is AI-powered HTS classification and duty recovery for licensed customs broker firms. Built on 220,000+ CBP CROSS rulings. Learn more →