Home / Guides / HTS Chapter Complexity Census

The US Tariff Schedule's Complexity Map: Which HS Chapters Hide the Most Lines (and the Most Money)

Ask five importers which part of the US tariff schedule is hardest to navigate and you will get five confident answers, most of them wrong. The honest way to settle the question is to count. Every chapter, heading and line in the Harmonized Tariff Schedule of the United States carries a machine-readable row, and once you tabulate them the "complexity map" of US tariffs stops being folklore and becomes arithmetic. This page is that arithmetic: a complete census of the current schedule, computed line by line from the USITC's published data rather than quoted from memory.

The headline structure looks like this. The schedule holds 3,362 eight-digit legal-rate lines spread across 92 chapters, plus another 19,831 ten-digit statistical rows that exist purely for reporting granularity. Those eight-digit lines are not distributed evenly: the ten busiest chapters hold 44.8% of them, while dozens of chapters manage fewer than twenty. Where the lines cluster, classification arguments, statistical errors and duty surprises cluster too.

Line counts are the schedule's population density

Counting rows sounds pedantic until you realize what a row is: a distinct combination of legal text and a legally binding rate. A chapter with 300-plus lines offers 300-plus chances to land on the wrong one, and each wrong landing has consequences that range from a lost preference claim to a customs penalty. The table below ranks every chapter that contains at least eighty eight-digit lines, alongside its ten-digit statistical footprint, how many of its lines enter duty-free, and the median pure ad-valorem rate among its non-free percentage-only lines.

*Median among the chapter's percentage-only eight-digit lines; — means the chapter has no such line (rates are compound or specific).

Two features of this ranking deserve attention before anything else. First, the machinery pair — chapter 84 and chapter 85 — jointly accounts for 503 eight-digit lines, 15.0% of everything the schedule taxes. Second, the apparel pair of chapters 61 and 62 adds another 371 lines, 11.0% of the total. Between them, four chapters cover more than a quarter of the entire tariff universe, which is why sourcing teams that only memorize "their" chapter keep getting ambushed by adjacent headings.

Free-rate density: the schedule's poles

Line counts measure complexity, but the mix of rates inside those lines measures pain. Here the census exposes a pole structure nobody warns newcomers about. At one extreme sits chapter 72, iron and steel: 109 of its 112 eight-digit lines enter the US duty-free, a 97% free-rate density, because decades of trade policy pushed raw steel into a zero-rated corner of the schedule. At the other extreme, chapter 61 — knitted apparel — contains exactly 1 free line out of 138, and chapter 62 manages just 3 free lines out of 233. Wearing something made of knit fabric is, statistically, one of the most reliable ways to owe duty entering the United States.

The same poles show up in the median rates. The pure ad-valorem median across chapter 61's percentage-only lines runs at 14.6%, against 4.2% in machinery chapter 84 and 2.7% in electronics chapter 85. Nothing in the harmonized system obliges chapters to behave this way; it is simply where decades of negotiated concessions left the furniture. But the practical consequence is real: an exporter moving from steel components to finished garments multiplies both their classification workload (more lines to choose among) and their expected duty bill (medians several times higher).

Sibling depth: when one heading is a whole menu

Chapters are the coarse layer of complexity. The fine-grained version lives inside individual six-digit headings, where the schedule hangs multiple eight-digit "siblings" with wildly different rates. The census finds 2,338 such families of eight-digit lines across the schedule; 1,776 of them contain exactly one line, while 226 headings carry three or more. The deepest family trees are genuinely bushy:

Depth alone is not the trap, though. The trap is when the siblings disagree about money. The single most instructive example in the schedule is heading 6204.69 — women's and girls' trousers of other textile materials. Its twelve statutory branches do not span a modest band: the lowest enters at 1.1% and the highest at 28.6%, a 27.5-point spread inside one heading, decided entirely by fiber content and construction details spelled out in the suffix descriptions.

Concretely: lines 6204.69.04, 6204.69.45 enter at 1.1%, while 6204.69.03, 6204.69.28 carry 28.6%. Same heading, same six digits, same day — 27.5 percentage points apart purely because Congress and USTR have negotiated fiber-and-construction carve-outs into this branch over decades. The suffix descriptions decide which one you owe, which is why apparel classification disputes rarely leave this heading's neighborhood.

Footwear's 6404.19 family tells the same story from a different angle: 22 eight-digit lines under one heading is the deepest branching in the schedule, spanning rates from 7.5% to 48.0% depending on whether the upper is textile, whether the shoe covers the ankle, and similar details that sound academic until a 40.5-point rate difference lands on the invoice. This is also why the schedule needs nearly twenty thousand statistical suffixes: the ten-digit layer exists so customs can see, statistically, what actually moves — and chapter 84 alone accounts for 1,706 of those suffixes while chapter 29, organic chemicals, claims 1,227 despite having only 89 legal lines.

Reading the complexity map strategically

If you classify products occasionally, the census condenses into three working rules. Rule one: know your chapter's line budget. If you operate in chapters 84, 62, 85, 52 or 61, you are working the five densest chapters in the schedule, and a "close enough" classification habit will eventually cost real money. Rule two: check the free-rate density before assuming a rate exists. In steel (chapter 72) or fish (chapter 03, where 79 of 83 eight-digit lines are Free), the default outcome is zero; in apparel the default is a double-digit percentage. Importers who transfer intuition across chapters in either direction systematically mis-price landed cost.

Rule three: treat deep headings as negotiation surfaces. When a single heading like 6204.69 spans 27.5 points, the difference between sibling lines is usually a fiber-content determination or a construction detail — exactly the facts a supplier can document for you if asked before shipment, and exactly the kind of finding CBP ruling disputes are made of. Pair this census with our guide to what the last digits of a ten-digit code actually do and the HS codes most disputed at CBP to see where complexity and controversy overlap.

Finally, keep the census's blind spots in view. Line counts measure structural complexity, not enforcement intensity: chapter 99's temporary provisions (Section 301 actions, safeguards, AD/CVD stacking) live outside this count and can multiply any number in the tables above. And statistical suffix counts reflect reporting granularity, not legal risk — although as anyone who has answered a Census Bureau query letter knows, the suffix layer generates its own paperwork weather. Use the map to decide where to slow down, not as a substitute for reading the actual line.

Frequently asked questions

What is the difference between an eight-digit HTS line and a ten-digit HTS code?

The first eight digits (two for the chapter, then product detail) carry the legally binding duty rate in column 1 of the Harmonized Tariff Schedule. The last two digits are statistical suffixes: they do not change the rate, but they decide which breakout your shipment is reported under, and misreporting them is an entry error you still have to fix.

Which HS chapter has the most eight-digit lines in the US tariff schedule?

Chapter 84 (nuclear reactors, boilers and machinery) leads with 343 eight-digit lines, followed by chapter 62 (woven apparel) at 233 and chapter 85 (electrical machinery) at 160. The top ten chapters alone hold about 45% of all eight-digit lines.

Do more subheadings mean higher duties?

Not directly, but the complexity leaders sit at opposite ends of the rate spectrum: roughly half of machinery chapter 84's eight-digit lines enter duty-free, while knit-apparel chapter 61 has exactly one free line out of 138. Complexity also concentrates negotiation exposure, anti-dumping coverage and reclassification risk wherever it concentrates lines.

How many statutory sibling lines can a single six-digit HS heading contain?

In the current US schedule, the deepest six-digit heading is footwear 6404.19 with 22 eight-digit lines, and 226 headings carry three or more siblings. One heading can therefore span rates from near-free to above 28%, as the worked example on this page shows.

Where can I check how complex my own HS chapter is?

Use the USITC Harmonized Tariff Schedule to count the lines under your candidate headings, then run each finalist through a duty calculator that applies both the general rate and any stacked trade-remedy additions before you commit to a classification.

Every figure on this page was computed from the USITC-published tariff file snapshot documented in the source strip below, and the generator refuses to build if the underlying data drifts from these anchors. Run your own numbers with the ExportSonar duty calculator, browse the full HS code directory, or compare plan levels on the pricing page.