What Changes When USITC Publishes a New HTSUS Revision
The U.S. International Trade Commission updates the HTSUS at least once a year, typically effective January 1, and most importers experience each revision as background noise — a new PDF, a new effective date, nothing that touches products already classified. That's the wrong mental model. An HTSUS revision doesn't just add or renumber provisions going forward. It can retroactively invalidate the legal basis for a classification made under the prior version, and most importers have no systematic way to know which of their SKUs just became wrong.
An HTSUS revision doesn't just add or renumber provisions — it silently invalidates every classification that depended on the old text, and most importers have no systematic way to know which of their SKUs just became wrong.
What USITC Actually Publishes, and How Often
USITC's tariff-schedule role is administrative and continuous: it compiles and maintains the HTSUS, incorporating changes to the international Harmonized System (which itself revises on a five-year cycle at the WCO level), amendments from U.S. trade legislation, new Chapter 99 provisions reflecting executive and trade remedy actions, and routine technical corrections. A new schedule typically takes effect January 1 each year, but Chapter 99 — the temporary and special-measure chapter — can change far more often than that, on whatever cadence new trade actions are announced.
The HTSUS itself is organized into 22 sections and 97 active chapters, with the first six digits of every code internationally standardized and digits 7 through 10 specific to the U.S. schedule. A revision can touch any part of that structure: heading text, section or chapter notes, subheading breakouts, or duty rates — and because GRI 1 gives notes the same legal weight as heading text, a notes change can shift a classification just as decisively as a heading rewrite would.
Why "New Version" Doesn't Mean "New Products Only"
Every classification is legally a determination made against a specific corpus version — the HTSUS text, notes, and structure in effect at the time. When USITC publishes a revision, it doesn't just change the rules for future imports. It can mean that a classification made in good faith six months ago now rests on heading text or notes that no longer exist in that form. The product hasn't changed. The law under it has.
This is the mechanism by which a classification silently goes stale. Nobody re-evaluates a product every time the schedule updates unless there's a system that flags it. In practice, most importers find out a prior classification is now wrong the same way they find most classification errors: an internal audit that happens to catch it, or CBP finding it first in a Focused Assessment — at which point it's an enforcement conversation instead of a routine correction.
The Corpus-Version Problem, Concretely
Treat the HTSUS as what it actually is: a version-controlled legal document, not a static reference. A classification pinned to "the HTSUS" without recording which specific annual version was in effect is missing the information needed to determine whether that classification is still valid after the next revision. Two classifications for the same product, made a year apart under two different corpus versions, may both have been correct at the time and still require different codes today — not because anyone made an error, but because the underlying text changed.
That's a materially different problem from the one Why the Same Product Gets Two Different HTS Codes from Two Different Brokers describes, where two people disagree about the same static law. Here, the law itself moved. Distinguishing the two matters, because the fix is different: a broker disagreement is resolved by recording and comparing traversals against one corpus version; a stale classification is resolved by tracking which corpus version every classification was actually made under, and re-evaluating against the current one on a defined schedule rather than waiting for an audit to surface it.
What This Means in Practice
The practical discipline a USITC revision demands is version-pinning: every classification record should state which specific HTSUS corpus version it was evaluated against, not just "the HTSUS" as if it were unchanging. When a new version takes effect, that record makes it possible to identify exactly which prior classifications touched a heading, note, or subheading that changed — instead of re-reviewing an entire catalog from scratch or, worse, not reviewing it at all until CBP does.
Why This Is a Software Problem, Not Just a Calendar Problem
Tracking USITC publication dates is a calendar task. Knowing which of your specific classifications are affected by what changed in a given revision is not — it requires comparing the current corpus against the specific version each classification was originally evaluated under, provision by provision.
Kanon monitors USITC publications and ingests each new HTSUS version as a discrete, immutable corpus version the moment it takes effect. Prior classification records stay pinned to the corpus version under which they were made, and new classifications are evaluated against the current one — so the question "which of our classifications does this revision actually touch" has a direct answer instead of requiring a manual re-audit of the entire catalog.
Frequently Asked Questions
Does every HTSUS revision affect every importer?
No. A given revision might only touch a handful of chapters or a specific set of Chapter 99 provisions. But without version-pinned classification records, there's no efficient way to know whether your specific products were touched without checking each one against the new text.
How often does Chapter 99 change compared to the rest of the HTSUS?
Much more often. Chapters 1–97 typically update on the annual January 1 cycle, but Chapter 99 changes whenever a new trade remedy action, exclusion, or executive order creates or modifies a temporary provision — which can happen multiple times within a single year.
Is a classification made under an old HTSUS version automatically wrong after a revision?
Not automatically — most classifications are unaffected by any given revision. But it can no longer be assumed correct without checking, because the specific heading, note, or subheading it relied on may have changed. The classification needs to be re-evaluated against the current corpus version, not just assumed to still hold.