By Andy Gaber · Published August 25, 2026 · Last updated August 25, 2026
Key stat: A CBP binding ruling under 19 CFR Part 177 is public — you can search cross-referenced First Sale rulings via CBP's CROSS database before filing your own, which is the strongest indicator of whether your fact pattern will clear.

You're probably paying duty on your middleman's profit margin, and nobody at your brokerage bothered to tell you there's a legal way to stop.
Here's the setup, and if you import through a trading company in Hong Kong, Shenzhen, or anywhere in Vietnam's supply chain, you already know this pattern cold. The factory makes the widget for $10. Your trading company buys it for $10, marks it up 20%, and invoices you $12. You pay CBP duty on $12.
That extra $2 isn't cost. It's markup. It's the trading company's commission for sourcing, quality control, sometimes just for existing between you and the factory. And at a 25% Section 301 rate, you're paying an extra 50 cents in tariff on every $2 of pure middleman profit. Multiply that across a few thousand units a year and you're bleeding real money on duty that has nothing to do with what the product actually cost to make.
There's a mechanism to fix this. It's called first-sale valuation. It's been legal since 1992, CBP has published guidance on it for decades, and most importers in the $1M-$20M range have never used it — not because it doesn't apply to them, but because nobody walked them through it. Your broker mentioned it once in passing. Maybe. This is that walkthrough.
I'll say this up front: I'm not a lawyer, and this isn't legal advice. First sale is real, it's well-established, and it's also exactly the kind of program where a bad setup gets you a CBP audit instead of savings. Read this to understand the mechanics and decide if it's worth pursuing — then get a customs attorney or licensed broker to actually build the program with you.
Most importers assume customs duty gets calculated on whatever they paid for the goods — the invoice from whoever sold it to them. That's usually correct, and it's called "transaction value" under 19 U.S.C. § 1401a, the statute that governs how CBP values imported merchandise for duty purposes.
But here's the piece almost nobody explains clearly: when there's a chain of sales before the goods hit U.S. soil — factory sells to a trading company, trading company sells to you — the law doesn't automatically default to using the last sale (the one to you) as the dutiable value. It can, under the right conditions, use the first sale in that chain: factory to middleman.
Think about the structure. There are two transactions happening before your goods clear customs:
Standard practice — what your broker does by default — is to declare the second sale price as the transaction value, because that's the invoice you're holding and paying against. First-sale valuation lets you instead declare the first sale price, provided you can prove that sale happened, was a real arm's-length transaction, and the goods were destined for the U.S. the whole time.
The dutiable base shrinks by exactly the middleman's markup. You're not hiding value, not underreporting anything — you're using a different, equally legal transaction in the chain as your customs value, one that CBP's own regulations and two decades of case law say you're entitled to use if you can document it.
This isn't a loophole in the "exploit a gap nobody noticed" sense. It's a deliberate, well-litigated feature of how transaction value works. It's just paperwork-heavy enough, and requires enough cooperation from a party who doesn't love opening their books, that most companies never bother.
The case that everyone in this space cites is Nissho Iwai American Corp. v. United States, 982 F.2d 505 (Fed. Cir. 1992). It's the Federal Circuit decision that cemented multi-tier transaction value as legitimate under U.S. customs law, and it's still the controlling precedent CBP and importers point to today.
Nissho Iwai involved exactly the structure described above — a manufacturer, a middleman, and a U.S. buyer — and the court laid out (building on earlier Customs Court precedent, principally E.C. McAfee Co. v. United States and the Treasury Department's own long-standing position) what's become the standard test. To use the first sale price as transaction value, you need to establish three things:
1. There was a bona fide sale between the factory and the middleman. Not a consignment arrangement, not an agency relationship where the "middleman" is really just acting on your behalf as your buying agent. An actual sale — title and risk of loss transferring from factory to middleman, with the middleman taking on the role of buyer and seller in its own right, not as your agent.
2. The merchandise was clearly destined for export to the United States at the time of that first sale. This is the piece that trips people up. It's not enough that the goods eventually ended up in the U.S. You need to show that when the factory sold to the middleman, everyone involved already knew — SKU, spec, and all — that this merchandise was headed to a U.S. buyer. Generic inventory the middleman bought and later happened to resell to you doesn't qualify. Purchase orders that name you, or reference your program, tied to that specific production run, are what establish this.
3. The price was set on an arm's-length basis, free of any non-market influences that would distort the value. CBP wants to see that the factory-to-middleman price reflects genuine market pricing — not a number manipulated to minimize duty, not a related-party transfer price set for tax reasons with no independent commercial logic. If factory and middleman are related parties, this element gets more scrutiny (more on that in the related-party section below), but related parties can still pass this test.
If you can document all three elements, you have a legitimate basis to declare the first sale price as your transaction value. CBP has recognized this framework in its own Informed Compliance Publication on Customs Value, which walks through multi-tiered transactions and cites Nissho Iwai directly as the governing standard. This isn't an aggressive interpretation of gray-area law — it's CBP's own stated position on how the statute works.
Let's make this concrete. All figures below are a hypothetical example — not a real importer's books — built to show how the arithmetic works.
Setup: You import $2M a year (at second-sale, middleman-invoiced value) of a product subject to a 25% combined duty rate (Section 301 plus base MFN rate, roughly representative of a lot of China-origin goods right now). Your Hong Kong trading company marks up the factory price by 20%.
First, back into the factory price. If $2M is the price after a 20% markup, the factory price is $2M ÷ 1.20 = $1.667M.
| | Value | Duty Base | Duty at 25% | |---|---|---|---| | Second sale (middleman → you) — status quo | $2,000,000 | $2,000,000 | $500,000 | | First sale (factory → middleman) — with program | $1,666,667 | $1,666,667 | $416,667 | | Annual savings | | $333,333 lower base | $83,333 |
That $83K figure is the headline case — a $2M importer with a 20% markup saves roughly $83,000 a year in duty, every year, for as long as the program stays intact and the markup stays roughly where it is. That's not a one-time win. It compounds.
Now the smaller, more common case — someone closer to your actual volume, maybe $600K a year in imports at the same 25% rate and a slightly leaner 15% markup:
| | Value | Duty Base | Duty at 25% | |---|---|---|---| | Second sale — status quo | $600,000 | $600,000 | $150,000 | | First sale — with program | $521,739 | $521,739 | $130,435 | | Annual savings | | $78,261 lower base | $19,565 |
To hit roughly $50K in annual savings — the number in this article's title — you need some combination of volume and markup that nets to about $200K of markup value stripped out of the dutiable base at a 25% rate ($200K × 25% = $50K). That's a $1.2M importer at a 20% markup, or a $1.5M importer with a leaner ~15% markup, or a $2.4M importer with the markup down around 9%. The mechanics scale linearly: bigger volume or fatter middleman margin, bigger savings. Thinner margin or smaller volume, and you need to weigh whether it's worth the compliance lift (see Section 9).
One more thing worth internalizing: this savings recurs every single year you keep the program compliant. Unlike a one-time drawback claim or exclusion request, first-sale valuation resets the baseline for every future entry. Set it up once, maintain it, and the $50K-$80K keeps showing up.
You need a genuine multi-tier supply chain — meaning there actually is a middleman between you and the factory, and that middleman is buying and reselling, not just brokering a deal on commission. If you already buy direct from the factory, there's no first sale to unbundle; you're already paying duty on the lowest legitimate number in the chain.
You also need the markup to be big enough to matter. A trading company charging 3-5% for logistics coordination isn't going to move the needle much once you account for the cost of running a compliance program. Markups in the 12-25%+ range — common with Hong Kong and Shenzhen trading companies serving as the interface between smaller Chinese factories and U.S. buyers who don't want to manage factory relationships directly — are where first sale earns its keep.
Unrelated-party middlemen are the cleanest case. Independent trading company, independent factory, arm's-length pricing on both legs almost by default because neither side has any incentive to distort the numbers for you. This is the easiest fact pattern to document and the one CBP scrutinizes least aggressively.
Related-party middlemen can absolutely work — plenty of first-sale programs run through a factory-owned or affiliate trading arm — but you're carrying a heavier documentation burden. CBP is going to look harder at whether that factory-to-middleman price reflects real market pricing or whether it's an internal number set for reasons that have nothing to do with what an unrelated buyer would pay. You'll need to affirmatively demonstrate the price wasn't influenced by the relationship — through comparison to prices charged to unrelated buyers for similar goods, or by showing the price reflects all costs plus a reasonable profit (the "test values" mechanisms built into the transaction value regulations). It's more work, not a dealbreaker.
Who doesn't qualify, or shouldn't bother: importers where the "middleman" is actually your buying agent (acting on your behalf, paid a commission, never taking title) — that's a different legal structure entirely (buying agency, not multi-tier sale) and first sale doesn't apply. Also skip it if your trading company is unwilling to disclose anything about the factory relationship — without their cooperation, you cannot build the file, full stop.
This is where first-sale programs live or die. CBP's Informed Compliance Publication on Customs Value is explicit that the burden of proof sits with the importer — you have to affirmatively demonstrate the three Nissho Iwai elements, not just assert them. Here's the file you need to build and keep current:
Missing any one of these doesn't automatically sink the program, but it weakens your file, and a CF-28 (covered in Section 7) will find the gap. Build the complete stack before you file your first F-flagged entry, not after.
Here's the unglamorous part: the biggest obstacle to first sale usually isn't CBP. It's your own trading company.
Approaching your middleman. You're asking them to hand you their factory invoice — the exact document that reveals their markup, the number they've spent years keeping opaque because it's the entire basis of their negotiating leverage with you. Expect resistance. Frame it correctly: you're not asking them to lower their price or cut their margin. You're asking them to let you pay U.S. customs duty on the factory number while they still get paid their full invoiced amount. Their revenue doesn't change. Your duty bill does.
Some will refuse outright, worried you'll use the factory price to renegotiate their markup down the road (a legitimate concern on their end, honestly — nothing stops you from doing that once you have the number). Some will cooperate if you frame it as a partnership: offer a longer-term purchase commitment, exclusivity on certain SKUs, or simply be straight that first sale is now table stakes for keeping the relationship at current volume. If your trading company is charging a large enough markup that first sale saves you tens of thousands a year, that's also a signal worth sitting with — it may be time to explore whether you can shift more of that value to a factory-direct relationship over time, first sale or not.
Broker setup. Your customs broker needs to be told explicitly that you're implementing first sale — this isn't automatic and isn't the default entry process. They'll need the documentation stack (Section 5) on file before they can properly flag entries, and most brokers will want a signed first-sale program agreement or memo from counsel before they'll put their name on the entries.
The "F" indicator. Since a reporting change that came out of the 2008 Farm Bill (the Food, Conservation, and Energy Act of 2008), CBP requires importers using first-sale valuation to flag it on the entry summary with an "F" first-sale indicator. This isn't optional cosmetic paperwork — it's how CBP tracks first-sale usage across the import population, and it means every first-sale entry you file is, by definition, self-identified to CBP as using this valuation method. That's not a reason to avoid it (flagging accurately is required either way), but it does mean sloppy first-sale entries are easy for CBP to find and query. Get the flag right, and get the file behind it right, before you file entry one.
Sequence in practice: build the documentation relationship with your middleman → get a customs attorney or experienced broker to review the fact pattern and confirm you clear the Nissho Iwai test → set up the file structure with your broker → file a test entry with the F indicator → monitor for any CBP inquiry → scale to full volume once the first entries clear clean.
A CBP Form 28, Request for Information, is CBP's standard tool for asking an importer to substantiate something on a specific entry before or after liquidation. First-sale entries draw CF-28s more often than standard entries, precisely because the F indicator makes them easy to flag for review, and because CBP knows first-sale files are exactly the kind of thing importers get sloppy about.
What CBP typically asks for on a first-sale CF-28:
The response playbook: Don't scramble. If your documentation stack was built correctly from day one, a CF-28 is a file-pull, not a fire drill — you already have everything requested, organized and cross-referenced. Respond within CBP's stated deadline (typically 30 days, though the form will specify), submit exactly what's asked for, and don't volunteer more than requested. If a CF-28 catches a genuine gap — a missing proof-of-payment document, a PO that doesn't cross-reference cleanly — that's the moment to loop in a customs attorney rather than freelancing an explanation. A weak response to a CF-28 escalates to a Notice of Action (CF-29) proposing a rate advance, or worse, a broader audit (Focused Assessment) of your import program. A tight, complete, prompt response usually closes the loop with no rate change.
The uncomfortable truth: a CF-28 on a first-sale entry isn't a sign you did something wrong. It's often just CBP doing routine verification on a valuation method they know gets used incorrectly a lot. Treat it as expected maintenance, not an emergency, provided your file is actually in order.
If your trading company and factory are related — same ownership group, parent-sub, whatever — you're navigating two federal agencies with different, occasionally conflicting incentives, and it's worth naming that tension explicitly.
Customs (CBP) wants to see a low, defensible transaction value at the first-sale tier, but insists that value reflect a genuine arm's-length price — not manipulated downward. CBP's whole concern with related parties is that you'll set an artificially low factory-to-middleman price specifically to shrink the dutiable base, which is exactly the kind of manipulation the "arm's-length" prong of Nissho Iwai exists to catch.
Tax authorities (IRS transfer pricing rules, and their counterparts in China/Vietnam) want to see a transfer price between related entities that allocates an appropriate amount of profit to each jurisdiction, generally under an arm's-length standard too — but the incentives there often push toward showing more value staying at the manufacturing/trading entity abroad (for a range of tax-planning reasons) rather than minimizing it.
Those two goals aren't automatically contradictory — arm's-length is arm's-length, in theory, whether you're proving it to CBP or the IRS. But in practice, companies that set transfer prices for tax reasons without thinking about the customs consequence often end up with a first-sale price that's hard to defend on either front: too high to help much with duty, or structured in a way that doesn't hold up as genuinely market-based when CBP asks for support.
If you're related to your factory or trading company, this is not a DIY section of the process. You need someone — ideally counsel or an advisor who understands both customs valuation and transfer pricing — looking at the whole picture together, not sequentially. CBP has an established mechanism (comparing to test values, or cost-plus analysis under the regulations) for related-party transaction value to hold up; it's just more paperwork and more scrutiny, not a closed door.
I'd rather tell you not to bother than watch you spend more on compliance than you save. First sale isn't free — it has real setup cost and real ongoing maintenance cost, and for some importers the math just doesn't clear.
Thin markups. If your middleman is only charging 5-8%, the duty savings on a modest import volume might be a few thousand dollars a year — not nothing, but maybe not worth the legal fees to set up a defensible program, the ongoing documentation burden, and the added CBP scrutiny (every F-flagged entry is a self-identified target). Run the math from Section 3 with your real numbers before committing.
Uncooperative suppliers. If your trading company won't share factory invoices, won't confirm the relationship structure, or actively resists — you don't have a program. You have a wish. Don't try to force a first-sale file together with incomplete documentation; a half-built file is worse than no program at all, because it invites a CF-28 you can't answer cleanly.
Volatile or ad hoc sourcing. If you're switching factories and trading companies every few months chasing price, the administrative overhead of building a new documentation file for every relationship may exceed what you save. First sale rewards stable, long-term supply chains where you can build the file once and maintain it, not spot-buying.
Low duty rates. If your product isn't hit with Section 301 or isn't otherwise carrying a high ad valorem rate — say you're at 2-4% MFN duty with no additional tariffs — the whole exercise nets out to a rounding error. First sale earns its keep specifically because current China-origin and certain other-origin tariff stacking has pushed effective rates into the 25%+ range for a lot of product categories. At low rates, skip it.
Cost of compliance exceeds savings, full stop. Legal fees to set up a program properly, the broker's added handling, the ongoing documentation maintenance, and the elevated audit risk all cost something. If your estimated annual savings (Section 3's math, run on your real numbers) don't clear that bar by a healthy margin — I'd want at least 3-5x the setup cost recovered in year one — it's not worth the operational complexity for a small operation.
Getting a first-sale program approved and running is not a one-time event. It's a living compliance obligation, and CBP expects the file to stay current.
Annual reviews. Prices change — factory costs shift, middleman markups get renegotiated, currency moves. Review your documentation file at least once a year to confirm the factory invoice, middleman invoice, and pricing still reflect current reality. Stale documentation on file is a liability, not an asset, if CBP ever pulls it.
Price-change controls. If the factory price changes mid-year, you need a process for capturing that change in your documentation before the next shipment, not reconstructing it after the fact during an audit. Build this into your PO process with the trading company: any price adjustment triggers an updated factory invoice on file, tied to the effective date.
What breaks programs — the real failure modes:
A straightforward sequence to figure out if this is worth pursuing, and if so, how to build it.
Do I need my supplier's factory invoice? Yes — it's the single most important document in the file. It's the direct evidence of the first-sale price, and without it, you have no basis to declare anything other than the price your middleman charged you. If your middleman won't provide it, you don't have a first-sale program, no matter how good the rest of your paperwork looks.
Will asking my trading company for this blow up the relationship? Maybe, if you handle it badly. Frame it as: their revenue doesn't change, you just want to pay duty on the factory number instead of theirs. Some will resist because they're worried you'll use the number to renegotiate later — a fair worry on their part. Have the conversation directly, and be honest that you might, at some point, ask about their margin again. Trying to sneak the documentation out sideways damages trust faster than asking straight.
Is first sale legal, or is it a loophole? It's legal — settled by Nissho Iwai American Corp. v. United States in 1992, and reflected in CBP's own published guidance on customs valuation since. It's not an aggressive interpretation of a gray area; it's a recognized valuation method with a clear three-part legal test. That said, "legal" doesn't mean "easy" — the documentation burden is real, and a poorly built first-sale file can still get you in trouble even though the underlying method is sound.
Does first sale work with Section 301 duties? Yes, and this is actually where it matters most right now. Section 301 duties are ad valorem — a percentage applied to the customs value. First sale reduces that customs value (the base the percentage applies to), so every dollar of markup you strip out of the dutiable base saves you 25 cents (or whatever your combined ad valorem rate is) directly. It doesn't reduce the rate — it reduces what the rate applies to.
What if my trading company and factory are related parties? It can still work, but you're carrying a heavier proof burden — CBP will want more support that the price wasn't distorted by the relationship, typically through comparison to unrelated-party pricing for similar goods or a cost-plus analysis. This is the scenario where I'd most strongly push you toward counsel rather than a DIY setup, because you're also navigating transfer-pricing considerations alongside customs valuation.
How much does it cost to set up a first-sale program? It varies with complexity, but budget for meaningful legal or broker consulting fees to get the fact pattern reviewed and the documentation structure built correctly the first time. Weigh that cost against Section 3's savings math — if your annual savings don't clear the setup cost by a healthy multiple, it's probably not worth pursuing yet.
Can I apply first sale retroactively to past entries? Generally, no — first sale needs to be declared and flagged (the F indicator) at time of entry. There are narrow post-entry correction mechanisms for other issues, but building a first-sale claim retroactively onto entries you already filed at second-sale value is a different, harder conversation to have with counsel, not something to assume works by default.
Does every product in my import mix need to go through first sale, or can I pick and choose? You can apply it selectively, by SKU, supplier, or product line. In fact, starting narrow — with your highest-volume, highest-markup, most stable supply chain — is the smart way to pilot the program before scaling it across your full catalog.
What happens if CBP disagrees with my first-sale valuation? They'll typically start with a CF-28 request for information, and if your response doesn't satisfy them, escalate to a CF-29 Notice of Action proposing a rate advance (essentially reverting you to second-sale value, plus potential penalties if they believe the claim was made in bad faith). A well-documented, good-faith first-sale file that gets challenged usually results in, at worst, a valuation dispute — not a fraud allegation. Sloppy documentation is what turns a valuation disagreement into something worse.
Is this worth it if I only import $500K a year? Depends entirely on your markup and duty rate. At a 20% markup and a 25% duty rate, $500K a year in imports could still save you around $17K annually (working through the math from Section 3) — worth it for some importers, marginal for others once you weigh compliance overhead. Run your own numbers using the worked example in Section 3 before deciding either way.