US duty drawback, explained for importers who have never filed

Duty drawback is a refund of certain duties, taxes, and fees paid on imported merchandise, available when that merchandise, or a substituted equivalent, is later exported or destroyed instead of being consumed in US commerce. The underlying logic, tracing back to one of the earliest acts of the US Congress, is straightforward: duty is meant to apply to goods that actually enter US commerce, not to goods that pass through the country on their way somewhere else or get built into something that ultimately leaves. If duty was paid on something that never really stayed here, drawback is the mechanism to get that money back.

It is one of the most underused refund provisions in US trade law, and mostly for a mundane reason: most importers never realize their own supply chain touches it at all, because the export or destruction that would trigger a claim often happens in a completely different part of the business than the import that paid the duty in the first place.

Unused merchandise drawback is the simplest category conceptually. The imported goods are exported or destroyed without being used in the United States, in essentially the same condition they arrived in, allowing for testing, inspection, and similar operations that stop short of actual use. Goods came in, never meaningfully entered US commerce, and went back out.

Manufacturing drawback covers imported materials or components used to manufacture an article in the United States, where that finished article is then exported. Duty paid on the imported input can be recovered even though the input itself was transformed along the way, as long as it can be shown to have gone into the exported product, either directly or, under the substitution rules described below, by way of an approved manufacturing ruling.

Rejected merchandise drawback covers goods that do not conform to sample or specification, arrive defective, or were shipped without the consignee's consent, and are subsequently returned or destroyed. This is the return-to-sender case: duty paid on goods that turned out to be the wrong thing and never should have stayed in US commerce in the first place.

Today's drawback program is set out in 19 CFR 190, and claims generally must be filed within five years of the date of importation of the merchandise the claim relies on. That is a long window by customs standards, and it is also why drawback opportunities pile up unnoticed: a shipment imported this year could still support a claim tied to an export that has not happened yet, or the reverse, a claim can reach back to cover imports from years earlier as long as both the export or destruction and the filing land inside that five-year clock.

Substitution is the feature that turns drawback from a niche, serial-number-matching exercise into something a company moving standardized product at real volume can actually operate. Rather than proving that the literal physical unit imported is the literal unit exported, which is often commercially impossible once goods are commingled in a warehouse or blended into a manufacturing run, substitution drawback allows a claim based on merchandise that is commercially interchangeable with the imported goods, matched at the same 8-digit HTSUS subheading, subject to further conditions depending on the drawback type. That is what makes the program workable for high-volume, fungible inventory instead of only for goods tracked one serial number at a time.

Most companies that could file drawback claims simply never do, for a few consistent reasons. The program sits at the intersection of import compliance, export operations, and inventory or manufacturing records, three functions that often live in different departments and rarely compare notes with each other. The paperwork burden is real: proving the link between an import and a later export or destruction takes organized records, not a rough recollection that some of that lot probably got exported eventually. And because the refund shows up well after the original duty was paid, it tends to lose the competition for attention against whatever transaction is happening today. The result is that drawback usually gets identified, if it ever does, during a compliance review or an outside audit, long after money that could have been recovered has quietly expired out of the five-year window.

What actually makes a claim provable comes down to records that tie three events together: the import entry, proving duty was paid and on what; the export or destruction, proving the goods, or a qualifying substitute, genuinely left US commerce or were destroyed under proper supervision; and, for manufacturing claims, the production records connecting the imported input to the finished, exported article. Inventory and bill-of-materials records that were never built with drawback in mind can often still support a claim, but only if someone goes back and reconstructs the linkage before the underlying records age out or the systems that held them get replaced. Because CBP can audit an approved drawback claim years after it was paid, the records that got a refund in the first place need to survive at least as long as the claim itself remains open to challenge. Treated as a live compliance program rather than a one-off recovery project, drawback is one of the few corners of US trade law where the incentive to keep good records and the incentive to get money back point in exactly the same direction.