UFLPA Compliance for Apparel Brands: What CBP Actually Requires

In this cluster

This page is for the brand founder or ops lead who has just been told their cotton programme has a UFLPA problem and cannot find a straight answer anywhere. CBP's forced-labor hub carries no apparel- or cotton-specific guidance — it publishes a generic document list that applies equally to polysilicon, tomatoes and t-shirts. This cluster pillar covers the legal framework and what CBP requires; the cost side sits next door in how to calculate landed cost for apparel imports.

What the Uyghur Forced Labor Prevention Act is and how it works through Section 307

The Uyghur Forced Labor Prevention Act is Public Law 117-78, signed 23 December 2021. Its operative provision — the rebuttable presumption — took effect 21 June 2022.

UFLPA did not create a new import prohibition. It created an evidentiary shortcut on top of one that already existed. The underlying ban is Section 307 of the Tariff Act of 1930, codified at 19 U.S.C. § 1307, which prohibits importing goods mined, produced or manufactured wholly or in part by convict, forced or indentured labor. Section 307 has been law since 1930; what changed in 2022 is who has to prove what.

Under a classic Section 307 action, CBP must develop information indicating forced labor before it acts. Under UFLPA, for goods meeting either trigger, CBP starts from the position that forced labor is present and you carry the burden of dislodging it. That reversal is why UFLPA reshaped cotton sourcing and Section 307 alone never did.

Two agencies matter. The Forced Labor Enforcement Task Force (FLETF) — chaired by the Department of Homeland Security, with USTR, Labor, State, Treasury, Justice and Commerce — owns the Entity List and the enforcement strategy. CBP enforces at the port. FLETF adds; CBP detains.

The terms people use for this, and which instrument each one means

Brands arrive at this subject through several different names for it, and they are not interchangeable. Two of them describe UFLPA triggers; the rest describe the older, broader ban underneath.

The table below maps each term to the instrument it actually refers to. The distinction decides which evidence you need and which clock you are on.

The practical consequence: screening your supply chain for a Xinjiang address answers only the first of these. The Entity List reaches companies anywhere in China, and Section 307 reaches producers anywhere on earth.

The two triggers for the UFLPA rebuttable presumption

The presumption attaches if either of the following is true. There is no third trigger, and there is no de minimis threshold below which a small amount of tainted input is tolerated.

Trigger one — XUAR nexus. The goods were mined, produced or manufactured wholly or in part in the Xinjiang Uyghur Autonomous Region. "In part" is the load-bearing phrase for apparel. A garment cut and sewn in Bangladesh from Indian fabric spun from a yarn lot that included Xinjiang cotton is within the presumption. So is a garment whose only Xinjiang content is the sewing thread.

Trigger two — Entity List. The goods were produced by, or sourced from, an entity on the UFLPA Entity List maintained by FLETF. This trigger is not geographic. A listed entity operating in Anhui or Shandong triggers the presumption exactly as a listed entity in Ürümqi does.

Both triggers run on inputs, not on the country stamped on the label. That distinction is the single most expensive misunderstanding in apparel compliance, and it is treated in full further down this page.

The UFLPA Entity List as of 3 August 2026: 187 entities and why geography-limited screening fails

The table below tracks the Entity List's size and the direction of enforcement. Counts are as of 16 August 2026; verify against the DHS entity page, which is the authoritative source.

Two details in that tranche matter more than the headline number.

First, 19 of the 43 new entities are located outside Xinjiang. If your screening process filters supplier addresses for XUAR provinces and clears everything else, it would have missed roughly 44% of the August 2026 additions. Geography is a starting filter, not a screen. The only defensible approach is to run every legal entity name in your supply chain — not just tier one — against the full list.

Second, textiles were a minority of the tranche but not an afterthought. Law-firm summaries of the additions (Kelley Drye's breakdown, which is a single-source count and should be treated as indicative) put roughly 23 in raw materials such as aluminum, carbon, gold and titanium, 9 in pharmaceuticals, 7 in food products, and 4 in textiles and apparel. Public reporting on the named textile entities differs between sources; check the DHS entity page rather than relying on any secondary list, including this one.

A worked example of screening scope

The reason "we screened our suppliers" is usually an inadequate answer is arithmetic. Take a mid-size brand running a cotton knit programme:

• 6 cut-and-sew factories
• each buying fabric from an average of 2 mills = 12 mills
• each mill buying yarn from an average of 2 spinners = 24 spinners
• each spinner buying lint from an average of 2 merchants or gins = 48 upstream counterparties

Total legal entities to screen: 6 + 12 + 24 + 48 = 90. Each of those 90 must be checked against 187 listed entities, and re-checked whenever the list moves — which, after an 18-month freeze, it just did by 43 in one day. CBP's own guidance states that "a gap for even one supplier can undermine the entire submission." A 90-entity supply chain screened 89 times is not 99% compliant; on CBP's stated standard it is a failed submission.

Most brands do not know all 90 names. Finding out is the work.

Apparel and cotton sit among CBP's twelve high-priority enforcement sectors

Appendix A of CBP's 2026 operational guidance names twelve high-priority enforcement sectors: aluminum, apparel, caustic soda, copper, cotton and cotton products, lithium, PVC, red dates (jujubes), seafood, silica-based products including polysilicon, steel, and tomatoes.

Apparel effectively appears twice — once as finished goods, once as the raw material. Cotton was one of the three original priority sectors named when the statute took effect, alongside polysilicon and tomatoes.

Be careful with detention numbers circulating online: third-party figures conflict by an order of magnitude, and we do not publish them. What is well supported is the rank. Electronics and solar have historically led detentions by volume, with apparel, footwear and textiles consistently second or third. Volumes fell sharply from January 2025 through mid-2026, and the July–August 2026 Entity List expansion signals a renewed enforcement wave. For counts and values, use CBP's enforcement statistics dashboard, linked in Sources, and note its as-of date.

Applicability review versus exception: the two paths importers conflate

This is the distinction that decides whether a detention is survivable. Most importers, and a lot of vendor content, use "rebut the presumption" to mean both. They are different processes with different burdens and radically different odds.

The table below compares the two routes out of a UFLPA detention. The practical asymmetry between them should drive how you build your file.

The exception route requires the importer to satisfy three statutory conditions simultaneously:

1. Full compliance with the FLETF UFLPA Strategy guidance and CBP's importer guidance;
2. Complete and substantive responses to all CBP requests for information; and
3. Demonstration by clear and convincing evidence that the goods were not produced wholly or in part by forced labor.

Then comes the part that kills it commercially. If the Commissioner grants an exception, they must report it to Congress within 30 days, identifying the goods and the evidence — and that report is public. No brand wants a public congressional filing naming it as the importer of goods with a conceded Xinjiang nexus. This reporting requirement, more than the evidentiary standard, is why exceptions are almost never the practical path.

The operational consequence: build your file so that you never need an exception. An applicability review is won on documents you can collect in advance. An exception is won on evidence that mostly does not exist and that you would not want published if it did.

What CBP's June 2026 Operational Guidance changed, and what it says will not satisfy a submission

In June 2026, CBP replaced its 2022 UFLPA-only guidance with a consolidated document. The specifics:

• Title: CBP Forced Labor Enforcement Operational Guidance for Importers
• Publication number: 5560-0526
• Document date: 9 June 2026, announced via CSMS # 68927213 on 12 June 2026
• Length: 89 pages, superseding the June 2022 guidance
• Scope: consolidates three authorities into one framework — 19 U.S.C. § 1307, UFLPA, and CAATSA § 321A (the separate rebuttable presumption covering goods made by North Korean nationals anywhere in the world)
• Contents: enforcement process maps for UFLPA, Withhold Release Order, Finding and CAATSA actions, plus appendices with supply-chain documentation recommendations and isotopic testing guidance
• Contact: ForcedLabor@cbp.dhs.gov

The consolidation is the substantive change. Before June 2026 an importer could reasonably treat UFLPA, WROs and CAATSA as three unrelated risks with three unrelated response playbooks. CBP now runs them as one enforcement architecture with four different clocks, which is exactly why the response timeline depends on how CBP codes your stop.

What CBP says will not satisfy a submission

This list is the most useful paragraph CBP has published for apparel importers, because every item on it describes something a brand is likely to send in first. CBP states that the following will not satisfy it:

• Affidavits standing alone
• Redacted records
• Untranslated documents
• Incomplete sub-tier supplier information

And the standard that governs all of it: "a gap for even one supplier can undermine the entire submission."

Read those bullets against how supplier documentation actually arrives. A mill sends a signed declaration that its cotton is not from Xinjiang — a standalone affidavit. A spinner sends a sales contract with the counterparty name and price blacked out as commercially sensitive — a redacted record. A gin sends its ticket in Gujarati or Mandarin — an untranslated document. A factory names its fabric supplier but cannot name the spinner — incomplete sub-tier information. Four of the most normal things in apparel sourcing are CBP's four stated failure modes.

Turning that into an actual bill of documents for a knit or woven garment is the job of the document-by-document cotton traceability file, which walks the eight-stage chain from farm to finished garment.

Why tariff origin does not protect you: 19 C.F.R. § 102.21 versus the forced-labor input trail

Textiles and apparel do not use the general substantial-transformation test for country of origin. They use the hierarchy in 19 C.F.R. § 102.21, applied in order: wholly obtained, then tariff shift, then the knit-to-shape rule, then the wholly-assembled rule, then most important assembly process, then last important assembly process.

The net effect for apparel is that the cut-and-sew country is almost always the country of origin. Chinese fabric cut and sewn in Bangladesh is Bangladeshi origin for marking and for duty.

This is the legal mechanism that made "China plus one" work, and it is exactly the gap UFLPA closes. Origin for tariff purposes does not launder the forced-labor provenance of the fibre. Your entry can be perfectly correct on origin, classification and valuation and still be detained, because admissibility runs on different rules.

The exposure concentrates where the cotton is imported. India and China are broadly cotton self-sufficient. Bangladesh and Vietnam are the world's two largest cotton importers — roughly 7.7 and 7.8 million 480-lb bales respectively in the 2025/26 USDA marketing year — which is why their fabric supply chains draw scrutiny despite neither country being a UFLPA target. If you are weighing origins on duty rate alone, read that alongside comparing India, Vietnam, Bangladesh and China as sourcing origins.

Why the end of de minimis pulled the DTC channel into forced-labor targeting

For years, direct-to-consumer apparel shipped in small parcels under the Section 321 de minimis exemption at $800 or less escaped the entry process entirely — no formal classification, no valuation review, no shipment-level admissibility screen. That channel is now closed.

• An interim final rule effective 24 June 2026, under 19 U.S.C. § 1321(b), indefinitely suspends the de minimis exemption for merchandise arriving via all modes other than the international postal network.
• The One Big Beautiful Bill Act, enacted 4 July 2025, permanently terminates de minimis effective 1 July 2027.

The compliance effect is larger than the duty effect. Every one of those parcels now carries an importer of record, an HTSUS number, a declared country of origin and a manufacturer identification — everything CBP's targeting systems need to run a forced-labor screen. A DTC brand that never thought about UFLPA because its average order value was $60 now sits inside the same enforcement perimeter as a container importer, usually without the compliance function. Who signs for that liability is the subject of who should be the importer of record; if the role itself is new to you, start with what an importer is and what the role requires.

How UFLPA connects to the 2026 Section 301 forced-labor tariffs

There is now a tariff-side companion to UFLPA. The two are separate and additive instruments, not alternatives.

Effective 12:01 a.m. ET on 24 July 2026, USTR imposed Section 301 tariffs on roughly 60 economies following determinations that 54 failed to impose forced-labor import prohibitions and 6 failed to enforce existing ones. Apparel is covered, at 10% for a lower tier including India, Bangladesh, Pakistan and Sri Lanka and 12.5% for an upper tier including China and Vietnam. Unlike the Section 122 bridge that preceded it, Section 301 carries no rate cap and no statutory time limit, so this regime is durable.

The point for compliance planning: your supplier country's forced-labor enforcement record now sets your duty rate, and your own supply chain's forced-labor exposure sets your admissibility. Paying the tariff does not buy admissibility, and clean traceability does not reduce the tariff. One knock-on effect worth checking this quarter — roughly doubled duty rates pushed many brands' continuous bonds into insufficiency, covered in customs bonds for apparel importers.

What to build now, and where the rest of this cluster goes

Three things are true simultaneously in August 2026: the Entity List grew by 30% after an 18-month freeze, CBP consolidated its enforcement architecture into a single 89-page framework, and the DTC channel that used to sit outside the entry process no longer does. All of that is a reason to have the file built before the detention, not after. The rest of this cluster:

• The document-by-document cotton traceability file — the eight-stage bill of documents from bale ID to finished garment, plus how isotopic testing actually fits.
• What to do when CBP detains your shipment — the 19 C.F.R. § 151.16 clock, four enforcement tracks with four different timelines, and the first 72 hours.
• Who should be the importer of record — where admissibility liability actually lands, and why a broker does not absorb it.
• Customs bonds for apparel importers — continuous versus single-transaction, and the bond-insufficiency wave the 2026 tariffs triggered.
• US apparel labeling and country-of-origin marking rules — the FTC rules that run alongside customs marking.
• Forced labor risk beyond Xinjiang — Withhold Release Orders, the June 2026 Jordan garment actions, and why duty-free trade-preference status offers no protection.

Frequently asked questions