.svg)
.svg){kind=icon}
Deferral decides when you pay. Status decides what you pay on. Most operations shop for the first and inherit the second.
Both tools stop the duty clock. That is the only thing they have in common. A U.S. foreign trade zone and a customs bonded warehouse sit under different statutes, permit different work on the goods, run on different calendars, and fix the duty rate at different moments. The last of those four differences is the one that moves landed cost, and it is the one most site-selection conversations never reach.
The gap between the two has widened in 2026, not narrowed. The broad duty layers now in force carry an admission restriction that removes the zone's most-cited advantage for covered goods. Meanwhile the bonded warehouse kept a feature nobody markets: the duty rate is not set until the goods leave. What follows is what each tool does under its own regulations, and the questions that decide which one belongs in your network.
Two tools, two statutes, two approval paths
A foreign trade zone exists under the Foreign-Trade Zones Act of 1934 (19 U.S.C. 81a-81u). Zones are licensed by the Foreign-Trade Zones Board under 15 CFR Part 400 and operated under CBP supervision under 19 CFR Part 146. Once a site is activated, it is treated for tariff and entry purposes as outside the customs territory of the United States. Foreign merchandise admitted there has not been entered. There is no consumption entry, no duty, and no quota charge unless and until the goods move into commerce.
A customs bonded warehouse exists under section 557 of the Tariff Act of 1930 (19 U.S.C. 1557). The facility classes and control requirements live in 19 CFR Part 19; the entry and withdrawal mechanics live in 19 CFR Part 144. Here the goods have been imported and entered, under a warehouse entry, and they sit in CBP custody with the duty unpaid.
That distinction is not academic bookkeeping. In a zone, the merchandise has not arrived, legally speaking. In a warehouse, it has arrived and the bill is outstanding. Every difference below descends from that one.
The approval paths differ accordingly. Bonding a warehouse is an application to CBP for a facility of a given class. Establishing a zone site, and especially obtaining authority to produce inside it, is a federal proceeding before the Board, followed by CBP activation of the physical space under 19 CFR 146.6. If your decision window is one quarter, that asymmetry alone may settle the question.
What each one gives you, side by side
| The question | Foreign trade zone (19 CFR Part 146) | Bonded warehouse (19 CFR Parts 19 and 144) |
|---|---|---|
| Legal position of the goods | Not entered; outside customs territory for tariff purposes | Entered for warehouse; in CBP custody, duty unpaid |
| How long they can stay | No outer limit; deferral runs until entry for consumption | Five years from the date of importation (19 CFR 144.5) |
| Duty if re-exported | None | None |
| Work permitted | Storage, exhibition, manipulation, and manufacture where the Board has granted production authority | Manipulation by class; manufacture only in bond and solely for export (Class 6), plus smelting and refining (Class 7) |
| When the duty rate is fixed | At admission under privileged foreign status, or at entry for consumption under nonprivileged foreign status | At withdrawal for consumption, always (19 CFR 141.69) |
| Dutiable base after processing | Price paid in the transaction that caused admission; zone labor, overhead, and profit excluded | Full value of the withdrawn article |
| Entry consolidation | Weekly entry available under 19 CFR 146.63(c) | Entry per withdrawal |
| State and local ad valorem tax | Exempt for admitted foreign merchandise (19 U.S.C. 81o(e)) | No equivalent exemption |
| Approval path | Board authorization plus CBP activation | CBP bond by class |
Three of those rows carry more weight than the rest. They are the next three sections.
The rate-setting moment is what decides your duty
Inside a zone, foreign merchandise carries one of two statuses, and the choice is the whole ballgame.
Under nonprivileged foreign status, merchandise is classified according to its character, condition, and quantity as constructively transferred to customs territory at the time the entry summary is filed, per 19 CFR 146.65(a)(2). If you assembled a finished article inside the zone from imported components, the rate that applies is the finished article's. The FTZ Board states the election directly in 15 CFR 400.1(c): the importer ordinarily has a choice of paying duty at the rate applicable to the foreign material as admitted, or, where it was used in production, at the rate applicable to the emerging product. When the component rate is higher than the finished-good rate, you elect the finished good. That is the inverted tariff, and it is the single benefit most often cited when someone pitches you a zone.
Under privileged foreign status, the arithmetic reverses. 19 CFR 146.41 fixes classification and rate as of the date the application is filed in complete and proper form. And 146.41(e) makes that status binding: it cannot be abandoned, and it follows the merchandise even after manipulation or manufacture changes its form. A locked rate is protection when rates are climbing. It is a ceiling you cannot escape when they fall.
The bonded warehouse offers no election at all, and that turns out to be its quiet advantage. 19 CFR 141.69(a) provides that merchandise entered for warehouse is dutiable at the rates in effect when the withdrawal for consumption is made. Nothing is locked in either direction. If a rate rises while the goods sit, you pay the higher figure. If a measure lapses, an exclusion is granted, or a court unwinds a duty layer, you pay the lower figure without filing anything to claim it.
Hold onto that asymmetry. The zone lets you pick a moment in the past. The warehouse hands you a moment you have not reached yet.
What work you can actually do inside each one
Zone production is where the value-added benefit lives, and it is routinely described wrong. There is no duty on value added in a zone. The opposite is true: under 19 CFR 146.65(b), the dutiable value is the price actually paid or payable in the transaction that caused the merchandise to be admitted. Labor, overhead, and profit generated inside the zone sit outside the dutiable base entirely. For an operation with meaningful U.S. content in its assembly, that exclusion often outweighs the rate election. It requires production authority from the Board, which is a filing, a comment period, and a wait.
The warehouse is narrower and more literal. 19 CFR 19.1 sets the classes. A Class 8 warehouse is established for cleaning, sorting, repacking, or otherwise changing the condition of imported merchandise, but not for manufacturing. Class 6 covers manufacture in bond solely for exportation. Class 7 covers smelting and refining of imported metal-bearing materials, for export or for domestic consumption.
Read those three lines against your own plan and the answer usually appears. If you intend to manufacture and sell into the U.S. market, the warehouse does not do it, and no amount of lease negotiation will change that. If you intend to relabel for retail compliance, build kits, repack bulk into consumer units, or hold inventory while a buyer decides, the warehouse does it well and you can be operating in weeks rather than quarters. And under 19 CFR 19.11, manipulation that changes the condition of the merchandise so that it takes a lower rate on withdrawal is not precluded. Narrower than the zone election, but not nothing.
What changed on July 24, 2026
The inverted tariff has been quietly closing for two years, and this summer it closed for most tariff-exposed goods.
Start with Section 232. Per the presidential proclamations, CBP has stated that any steel or aluminum article subject to Section 232 duties, other than those eligible for admission under domestic status, that is admitted into a U.S. foreign trade zone must be admitted as privileged foreign status, and will be subject on entry for consumption to the ad valorem rates tied to its classification. Manufacture in the zone does not by itself pull an article into Section 232, but the status is binding once granted.
Then came the broader layer. On July 23, 2026, USTR issued final action in sixty Section 301 investigations concerning the failure of those economies to impose and effectively enforce a prohibition on imports produced with forced labor. The duties took effect at 12:01 a.m. eastern time on July 24, 2026, at 10 percent for one group of economies, including Canada, Mexico, India, Indonesia, Malaysia, and the United Kingdom, and 12.5 percent for the rest, with product exemptions listed in the annexes and with articles already subject to Section 232 tariffs carved out. The notice of action then adds the operative line for anyone running a zone: a covered product admitted into a U.S. foreign trade zone, except one eligible for admission under domestic status, only may be admitted as privileged foreign status as of the date the additional duty is imposed.
Put the two together and the practical picture is this. For non-exempt goods from a covered economy, the zone election is unavailable. The rate is fixed on admission. What survives is real and still worth money: no duty on re-exports, the exclusion of zone value added from the dutiable base, weekly entry, and the state and local tax exemption. What does not survive is the promise of converting a high component rate into a lower finished-good rate.
The same notice points, almost incidentally, at the other tool. Its duties apply to goods entered for consumption, or withdrawn from warehouse for consumption, on or after the effective date. Warehoused merchandise pays what the schedule says on the day it leaves. In a year when one duty regime was invalidated in February and its replacement expired by operation of law in July, that is not a technicality. It is optionality, and it costs nothing to hold.
One honest caveat. These layers move, and the Section 301 action is already being contested. The structural difference between the two tools does not move. Build the decision on the structure, and revisit the layer every quarter.
Four questions that settle it
Where does the merchandise end up? If most of your volume is re-exported, both tools eliminate the duty and the warehouse usually gets you there faster and cheaper. If most of it enters U.S. commerce, keep reading.
Do you transform the goods, and for which market? Manufacture for domestic sale points to a zone with production authority, or to neither. Repacking, relabeling, kitting, and condition changes point to a Class 8 warehouse.
Is your exposure a rate you want to lock, or a rate you expect to move? Where privileged foreign status is mandatory anyway and rates are trending up, the zone lock has value. Where you are waiting on an exemption, an annex revision, or a litigation outcome, the warehouse holds the option.
How many entries do you file, and how long will the goods sit? Weekly entry under 19 CFR 146.63(c) consolidates a week of removals into one entry, and because the merchandise processing fee is assessed per entry with a fiscal-year ceiling, currently 0.3464 percent of value with a maximum of $651.50 per entry per CBP, high-frequency operations recover real money on fees alone. Against that, remember the warehouse clock: five years from importation, with no extension.
A distributor we worked with had modeled a zone on the strength of the inverted tariff, on a line where the component rate ran above the finished-good rate. Once the admission restriction applied to their origin, the election disappeared and the model lost most of its return. Their actual profile, roughly forty percent re-export to Latin America, sixty percent domestic release across unpredictable windows, and no transformation beyond repacking, pointed somewhere else entirely: a bonded facility, operating inside a quarter, with the duty on the domestic portion set on the day each order shipped. Same deferral, a fraction of the setup, and the rate exposure pointed in the direction they actually expected rates to move.
What we would look at first
We run brokerage on both sides of this border, with three Patentes Nacionales in Mexico and a U.S. Corporate Customs Brokerage License, across more than 190,000 customs operations a year at 39-plus ports, and we hold more than 600,000 square feet of bonded warehouse space in that network. The pattern across those operations is consistent. The operations that get this decision right start from their own flow data, the export-to-domestic split, the dwell distribution, the entry frequency, and the origin mix, and only then look at facilities. The ones that get it wrong start from a facility someone offered them and work backwards to justify it.
If you are weighing the two, the first artifact to build is not a lease comparison. It is a twelve-month picture of where your goods actually went, how long they waited, and which duty layers touched them. That document answers the question faster than any site tour.
Talk to a Joffroy expert about modeling your zone-versus-bonded decision against your own entry and dwell data, and about the admission-status exposure on your current origin mix.
Deferral answers when you pay. Status answers what you pay on. Choose the tool that gets the second question right, because the first one is the easy half.
TRADE. UNDER CONTROL.
