Vietnam dried fruit EU import requirements: which limits apply

Responsible fruit sourcing on a Vietnamese dried line is an arithmetic problem and a filing problem. An EU maximum residue level is set on fruit as harvested; drying removes 70 to 90 per cent of the water and concentrates what stays dissolved in it. Under Regulation (EU) 2023/915, where no dried entry exists, that arithmetic falls on the food business operator.

Does a residue result on fresh fruit prove the dried product complies?

No. Under Regulation (EC) 396/2005 a maximum residue level attaches to the commodity as defined in Annex I, which for most fruit is the whole product including peel once the crown or stalk is off. Commission Implementing Regulation (EU) 752/2014 carries those product definitions; banana, code 0163020, is defined as the whole product after removal of the crown. That is the number a laboratory reports against when fruit is sampled at the farm gate, and it is not the number that governs the goods in the carton.

Article 20 of the same regulation sets the mechanism for processed food, with the implementing guidance at SANTE/10704/2021 Rev 1, in force since 19 March 2025. The limit for a processed product is the limit for the corresponding unprocessed product multiplied by a processing factor: the ratio of residue concentration after processing to residue concentration before it.

Most buyers who have met a processing factor met the peeling case, where it runs well below one because post-harvest treatments sit on the skin that gets discarded. Drying runs the other way.

A lot of fresh fruit sitting comfortably at 60 per cent of its limit can produce a dried product several times over it, and the dried product is what has to comply.

How many kilograms of fresh fruit made one kilogram of this product?

That is the question, and it is answerable in one sentence by anyone who has run the line. The concentration factor is the water-loss ratio divided by whatever the process adds back, and only the facility’s mass balance gives it. A freeze-dried piece is fruit and nothing else: FD Food, a Vietnamese producer, publishes 30 g of finished freeze-dried jackfruit from roughly 150 g of peeled and de-seeded fresh fruit, about five to one on an edible-bulb basis, and it states the basis, which is the only reason the figure is usable.

A soft-dried piece is a different object. The published specifications on this site’s own soft-dried mango and soft-dried jackfruit carry cane sugar to at least 70 °Bx and at least 68 °Bx respectively, so a large amount of contaminant-free mass has been added back after the water left. A vacuum-fried chip declared by Vinamit as 98 per cent fresh jackfruit plus vegetable oil has had a smaller amount added. Sugar and oil dilute.

What is added also has to reach the label. Sulphur dioxide and sulphites must be declared as an allergen above 10 mg/kg or 10 mg/L under Annex II of Regulation (EU) 1169/2011; the permitted use levels for dried fruit sit in Annex II of Regulation (EC) 1333/2008 and have to be read there, not assumed. Sulphur fumigation is standard practice in Vietnamese domestic drying, and a domestic drying-shed process does not transfer to an export container without that check. If an anti-browning or colour-holding dip has been used it belongs in the ingredient statement, and a product carrying a dip is not a 100 per cent fruit product.

Who has to prove the processing factor, and where do you find one?

The food business operator. SANTE/10704/2021 Rev 1 is explicit on three points that decide who does the work: whether a processing factor is applied is a decision for the competent authority in the importing country, not a choice the business makes for itself; responsibility and due diligence rest entirely with the food business operator; and where an available factor does not settle the question, the authority may require the operator to produce either evidence that the incoming raw material met the limit, or a processing factor generated for its own process.

The EU database of processing factors, second update published in 2024 as EFSA Supporting Publications EN-8738 by Kittelmann and colleagues, is searchable by active substance and processing operation, and the pair a Vietnamese dryer needs is frequently absent from it. A factor belongs to a triplet: this active, this fruit, this process. The published median of 0.07 for imazalil on peeled citrus is a real figure, and carrying it into a jackfruit or mango file is the first mistake on the list.

Nor is it safe to assume that processing always helps. In the United States, 40 CFR 180.507 sets azoxystrobin on whole banana at 2.0 ppm and on banana pulp at 0.1 ppm, twenty times stricter. Regulators have in places already set the flesh limit below the commodity limit, so each active in the spray programme has to be checked for its own sub-entry rather than assumed into the general case.

Generating a factor where none exists has a minimum design: three independent lots, two replicates each, paired samples from the same lot, an ISO/IEC 17025 laboratory with the active inside its accredited scope, limit of quantification and measurement uncertainty reported, and the trial run on the commercial process rather than a bench imitation.

Which EU limits apply to a dried fruit when the regulation has no dried entry?

The contaminants instrument is Regulation (EU) 2023/915 of 25 April 2023, which replaced Regulation (EC) 1881/2006. The older number is still quoted on supplier pages and in trade guidance, and which one a supplier cites is a fast test of how recently they read anything. Lead in fruit generally sits at 0.10 mg/kg.

For a dried tropical fruit there may be no separate entry at all. No cadmium maximum level for jackfruit or for dried mango was established for this article, and printing a plausible figure would be worse than saying so. What the regulation supplies instead is the rule that answers the question: where a food is dried, diluted or otherwise processed and no specific maximum level is fixed for it, the operator applying the annex must take account of the concentration or dilution factor and of the effect of processing. Absence of a dried line in the annex is not absence of obligation.

Contaminants run in the opposite direction to pesticides, and the mechanism is the reason. Lead and cadmium do not arrive on the skin. They come up from the soil through the root, so peeling, which rescues the pesticide arithmetic, does nothing for the metal arithmetic, while every operation that removes water concentrates them. Metal control happens in the choice of growing area, which is where responsible fruit sourcing across a mixed grower base starts. Cadmium in frozen pineapple from Costa Rica is the recorded precedent that this is not theoretical for processed tropical fruit.

Chlorate is the exception that tells you where to audit. CBI reports 0.3 mg/kg for tropical fruits including banana, mango, papaya, pineapple, passion fruit, durian and soursop, citing Regulation (EU) 2020/749, with perchlorate under Regulation (EU) 2020/685. Chlorate does not come from the field. It comes from chlorinated wash water inside the plant, which makes it the one contaminant a facility creates for itself, and the one a buyer audits at the facility rather than at the farm.

Two lines are worth not paying for on a default panel. Patulin limits, 50 µg/kg in the EU and in Codex, apply to apple rather than to tropical fruit. No aflatoxin or ochratoxin A limit specific to these dried tropical fruits was located for this article, and the well-known dried-vine-fruit and dried-fig entries do not generalise to dried mango or jackfruit; the mycotoxin section of Annex I to 2023/915 is where a buyer settles it.

Compare how China solved the same problem. GB 2762 sets a numeric cadmium limit of 0.05 mg/kg, and since 1 December 2024 every consignment of durian and jackfruit is tested through a GACC-recognised laboratory, with lots that fail or arrive without the certificate returned or destroyed. Neither regime is laxer. They place the burden in different places: the Chinese requirement is a scheduling problem, the European one a documentation problem.

Is Vietnamese dried fruit subject to increased official controls at the EU border?

The four Vietnamese entries on the list under Regulation (EU) 2019/1793 are fresh-produce lines. Regulation (EU) 2026/194 of 28 January 2026, applying from 18 February 2026, amended that list and added nothing for Vietnam. Work from the consolidated version rather than from any single amending act.

Nothing found for this article places a Vietnamese dried product on that list. That is a search that did not find something, not a verified exemption, and the distinction matters for a precise reason: those annexes are written against CN codes. This company sells dried okra chips while fresh Vietnamese okra sits at 50 per cent. Whether the dried form falls inside or outside that entry is decided by the code the goods travel under, which is the same question as the tariff heading, and why the process rather than the fruit decides the heading is worth settling before a first shipment. Classification and border control are one problem, not two.

Know the difference between the two regimes. An Annex I listing raises the frequency of checks at the border and holds lots for days. An Annex II listing additionally requires an official health certificate and analytical results from a recognised laboratory with every consignment, produced before export. Nông Thôn Việt reported in March 2026 that analysis runs at 600 to 1,000 euros per sample, with a container delayed a week attracting demurrage of up to 3,000 euros.

Who holds each certificate, and who cannot hold one at all?

Sort the file into four hands before booking an audit. A badge row hides the only question that decides where the audit goes: for each document, who is the legal holder, and what happens to it when that party changes.

The exporter’s row certifies nothing about a facility or a field, and that is the audit map rather than a weakness. The correct construction is the one this site’s product pages already use: made at FSSC 22000- and HACCP-certified partner facilities.

The layer buyers most often assume belongs to the supplier is the grower’s. GLOBALG.A.P. is held at farm level, and so is the spray-programme record: actives used, number of applications, date of last application, pre-harvest interval. That record is the only document that makes a residue result predictable rather than a lottery. VinFruits Global requires the growing area’s spray-programme declaration before raw material is accepted, because a residue result without the programme behind it describes one lot rather than a supply. Certified area is thin: under 2,000 hectares of Vietnamese mango carried VietGAP or GlobalGAP in 2025 against roughly 87,000 to 115,000 hectares planted, so a certified-farm claim is a claim about one linked growing area, not about a country.

One layer does not apply here at all. Planting-area codes and packing-facility codes attach to a parcel of land and to a building, never to a trading company, and they are a China-side requirement rather than an EU one. A processed line needs none of them: dried goods, purée, concentrate and IQF need no protocol and no code, only the facility’s food-safety file. Where the codes do apply they are weak as a standing assurance: 403 planting-area codes and 240 packing-facility codes were warned or suspended across 2025 and 2026, and active packing-facility codes fell by around 36 per cent in thirteen months. A certificate is a statement about a date, which is why the working rule is to re-verify against the issuing register before every lot rather than file a scan.

Why an internal limit set level with the MRL is not a limit

Because official enforcement carries a buffer that a commercial contract does not.

Under SANTE/11312/2021, EU official control of pesticide residues applies a default expanded measurement uncertainty of 50 per cent, coverage factor k = 2 at 95 per cent confidence, and concludes that a maximum residue level has been exceeded only where the measured value minus that expanded uncertainty is above the limit. A result of 1.4 mg/kg against an MRL of 1.0 mg/kg is therefore not an official exceedance, because 1.4 less 50 per cent is 0.7. The mirror case is the one that costs a shipment: a result of 0.9 mg/kg against the same 1.0 mg/kg limit may genuinely be 1.35 mg/kg, so a lot sitting just under the line is not compliant, it is unmeasured. An internal specification for a dried line therefore has to sit below the legal number, and incoming raw material has to be bought against a threshold near the finished-goods limit divided by the concentration factor, with a further margin subtracted.

Buyer and retail-chain limits sit below the legal one, sometimes as a percentage of the MRL, sometimes as a cap on the number of actives detectable at all, and a contractual limit carries no uncertainty buffer, because that buffer belongs to official enforcement. Where an official control result is disputed, Article 35 of Regulation (EU) 2017/625 gives the operator the right to a second expert opinion at its own cost, including re-analysis of the remaining sample, except where re-analysis is impossible or irrelevant. That mechanism is worth mirroring in the commercial contract, because the private dispute has no statutory version of it.

What does responsible fruit sourcing look like when the data is unflattering?

It looks like printing the numbers. Listings and national measures are earned by monitoring data, and a sourcing page that omits the data is a brochure. French monitoring data for 2024, published by DGAL in February 2026, found 15 of 47 samples of Vietnamese passion fruit over MRL, 32 per cent, with omethoate named, against under 5 per cent for Thailand and Ecuador; Vietnamese dragon fruit in the same exercise was 1 of 9. Around 30 per cent of durian and jackfruit samples in the Mekong Delta exceeded the cadmium threshold, reported by Thương hiệu & Công luận in May 2026.

Two counts of RASFF notifications against Vietnamese goods for 2024 are in circulation and they do not agree. One records 114 notifications, 2.2 per cent of 5,268 in total, with chemical residues the largest cause group at 31.3 per cent, and 16 of 624 in the first two months of 2025. The other, reported by Tuổi Trẻ on 27 March 2026 from the Ministry of Agriculture and Environment, records 64 in 2024 falling to 17 in 2025. Neither has been checked against the RASFF Window itself. The boundary is almost certainly scope: all Vietnamese food and feed against agri-food of plant origin only, or notifications against border rejections. Until the portal is queried directly neither displaces the other, and a supplier quoting the lower one without saying why is quoting a number they have not checked.

The counter-evidence points at the processed line. Across 2021 to 2025 there were 15 RASFF notifications on fresh passion fruit for pesticide residues and none at all on passion fruit juice, because pressing discards the peel where most residue sits and bulk product is blended across lots, which flattens peaks. The caveat travels with it: blending flattens a residue peak and spreads a contaminant evenly through the whole batch.

One divergence has to be checked before any lot ships. EU-wide law fixes an MRL, a threshold a consignment can sit under. France issued a decree on 5 January 2026 prohibiting imports of non-EU food carrying residues of mancozeb, thiophanate-methyl, glufosinate, carbendazim or benomyl at all, a binary rather than a threshold, effective from around mid-February 2026 for one year or until an EU-wide measure replaces it. Carbendazim is a common fungicide on tropical fruit, so a consignment comfortably compliant EU-wide can be refused entry at one Member State, and because the measure is time-limited its current status has to be confirmed before anyone relies on it either way.

Which clauses make any of this enforceable?

A threshold with no test method, no sample size, no sample state and no named party to measure it is unenforceable in a dispute. Six clauses do the work on a dried line, and all six paste into a specification annexe.

• A spray-programme declaration from the growing area: actives, number of applications, date of last application, pre-harvest interval. No declaration, no raw material accepted.
• A full multi-residue panel before the first lot from any new growing area or facility.
• An internal limit stricter than the MRL, which is not optional on dried and concentrated goods.
• The right to sample raw material at the factory weighbridge, not only finished goods on the pallet. A finished-goods result arrives too late to do anything but reject.
• A supplier undertaking to notify any change of active substance in the programme.
• An explicit allocation of who pays laboratory, storage, re-export and destruction costs when a lot fails.

Name the method beside every threshold: Salmonella to ISO 6579-1, Listeria to ISO 11290-1 and -2, a multi-residue panel by LC-MS/MS and GC-MS/MS with proficiency testing on the fruit matrix, heavy metals by ICP-MS. Ask the laboratory for the accredited scope entry naming this analyte, this method, this matrix and this limit of quantification, because ISO/IEC 17025 accredits each test on each matrix rather than the laboratory as a whole.

Three fields of an arbitration-laboratory clause are habitually forgotten: a sealed retained sample signed by both parties at handover rather than a sample taken once the dispute has started; the storage condition of that sample, which has to be the condition printed on the label; and the comparison rule, whether exceeds means the measured value above the limit or the measured value minus the expanded uncertainty above it. The third is where the better-prepared party wins.

The origin file is the traceability file. Under EVFTA the product-specific rule for Chapter 8, where dried fruit sits at heading 0813, turns on the fruit being wholly obtained, grown and harvested in Vietnam, so there is no route for third-country fruit into a Chapter 8 line claiming EU preference. The failure mode is not fraud, it is warehousing: imported and domestic fruit sharing one store, one shift and one set of books, so that a post-clearance verification asking which raw material went into a given lot gets the answer that it cannot be separated, and that answer carries the same consequence as a false declaration. Verification runs through REX, and the loss that hurts is deregistration rather than the penalty. Ask to see the raw-material store ledger separated by origin. It is the cheapest test of whether a traceability claim is a system or a slogan.

Common questions

Does a residue test on fresh fruit prove the dried product complies?

No. An EU maximum residue level under Regulation (EC) 396/2005 is set on the commodity as defined in Annex I, normally the whole fruit including peel after the crown or stalk is removed. Drying removes most of the water and concentrates whatever was dissolved in it, so a fresh result at a comfortable fraction of the limit can become a finished-goods result above it. The dried product is what has to comply.

Which EU limits apply to dried fruit when the regulation has no dried-fruit entry?

Regulation (EU) 2023/915 of 25 April 2023, which replaced Regulation (EC) 1881/2006, carries a rule for foods that are dried, diluted or otherwise processed and for which no specific maximum level is fixed: the operator applying the annex must account for the concentration or dilution factor and for the effect of processing. The absence of a dried entry is not an absence of obligation. It moves the arithmetic onto the operator.

Is Vietnamese dried fruit subject to increased official controls at the EU border?

The Vietnamese entries on the list under Regulation (EU) 2019/1793, as reported at 27 March 2026, are dragon fruit at 30 per cent, chilli at 50, okra at 50 and durian at 20, and all four are fresh lines. Nothing found here places a Vietnamese dried product on the list, but that is a search that found nothing rather than a confirmed exemption. Those annexes are written against CN codes, so check the consolidated version against the code your goods travel under.

Which certificates can the exporter provide, and which only the factory or the grower?

An exporter signs commercial documents: contract, invoice, packing list, bill of lading, the lot specification, the certificate of analysis as issued, and the allergen and additive declaration. A GFSI scheme certificate such as FSSC 22000, BRCGS, IFS or SQF is held by the facility and cannot be transferred to a trading company. GLOBALG.A.P. and the spray-programme record are held at farm level by the grower.

What should a dried-fruit COA contain, and what does a missing line mean?

A certificate of analysis states which parameters were measured, on which sample, with which results. Anything absent from it is something nobody measured. For a dried line, ask for moisture, water activity, microbiology, a multi-residue panel, SO2 and additives, each with the lot code matching the goods received, the sampling and test dates, the method reference, the limit of quantification and the measurement uncertainty.