Pre-Shipment Inspection of Frozen Vegetables: What to Sample and Record

A pre-shipment inspection of frozen vegetables should establish four things: which goods were presented, how samples were selected, what was checked, and who decided whether the goods could proceed. Agree those points before the visit. A collection of product photographs and a final “pass” is difficult to interpret if the lot, sampling basis or unresolved tests are missing.

Start with the order specification and an identified shipment breakdown. Separate physical quality and packing checks from temperature assessment and laboratory testing, because each has its own sampling and reporting requirements. Keep the inspected lot linked to its sample units, findings, supporting records and release status, then confirm that the same accepted material enters the loading handoff.

We coordinate that inspection scope with the buyer, partner facility and agreed inspection party. We agree who will complete pending work and how its outcome will reach the person making the order decision. Inspection findings apply to their stated scope; they do not demonstrate that every unit was examined or replace the relevant food safety controls.

Define the shipment and lots before booking inspection

Send the inspector a brief that identifies the product, form, pack, order quantity, destination requirements, specification revision and approved sample reference where applicable. Include the planned production or packing completion, inspection window and loading date. A visit arranged before the required goods or records are available may answer only part of the intended question.

List the production lots or other agreed inspection groupings within the shipment. Record the carton or pallet quantities and how the stock can be located. A container booking is a logistics reference; it does not by itself establish that all its proposed contents were produced under the same conditions. If different products, forms, production lots or packing versions are involved, make that visible before sampling begins.

Codex’s General Guidelines on Sampling, CXG 50-2004, revised in 2023, explains lot characteristics, sampling plans and the handling of non-homogeneous material. Its guidance is not a ready-made sampling instruction for every export order. Check whether the proposed grouping and selection method suit the property being assessed.

For example, imagine an order containing broccoli florets from two production lots and spinach portions from a third. Keep each identity and its quantity available to the inspector. Do not present a single hand-picked bowl as the representation of all three. This example illustrates the identification problem; it does not prescribe three samples, a particular inspection level or a rule for combining results.

Confirm access to the stock, suitable cold conditions for the work and the documents needed at the visit. If some goods are inaccessible, still being packed or absent, the report should state the quantity actually presented and the limitation. Decide whether that coverage meets the inspection brief instead of allowing the report heading to imply that the full order was examined.

Select samples for each inspection purpose

State the inspection unit, selection method, sample quantity and decision rule for each required check. A carton selected for pack-mark verification is not automatically the laboratory sample for a microbiological test. A portion opened for a size assessment may no longer be suitable for another analysis that requires protected collection.

For acceptance sampling, the number selected and the acceptance criteria belong together. Codex distinguishes random selection from convenience sampling and explains that sampling retains risks of accepting or rejecting a lot incorrectly. Opening only the easiest cartons to reach is not equivalent to a documented random selection across the defined stock.

Ask the responsible inspector or laboratory to specify the plan appropriate to the product, inspection purpose and applicable requirements. If an inspection level or AQL is required, identify the agreed standard and the complete decision basis. An AQL value on its own does not tell a warehouse operator which units to select or a buyer what the reported conclusion covers.

Define the product form before defining a defect. Compact spinach portions may be the ordered format, whereas unwanted joined pieces may matter in a product ordered as loose, free-flowing pieces. Record what should be separated and how it is assessed under the agreed method. Breaking an intentional portion apart and classifying the resulting fragments against a loose-leaf requirement would answer the wrong question.

Keep each selected unit identifiable during the inspection. Where separate units must be tested separately, preserve that separation. Where compositing is part of an agreed laboratory method, the laboratory plan should define it. Extra targeted samples taken because a damaged carton or unusual area needs investigation should be labelled as such, with their location and reason, so they are not confused with the planned selection.

Record physical quality against the product form

Inspect the fields that affect the ordered use. A broccoli floret specification may need crown size, attached stem, loose stem pieces and fines distinguished. A diced vegetable specification may need dimensions, off-size pieces and broken material. A leafy portion may need a defined portion form and preparation-based assessment. The same word, such as broken, can describe different observations across these products.

For broccoli, the crown and its supporting stem form one connected piece. Define the measurement points and the treatment of separate stem material before reporting a percentage. Our guide to broccoli floret size and stem ratio explains that particular distinction. In the pre-shipment report, identify the method used and the sampled material to which the result applies.

Give every calculated result a denominator. “Four broken pieces” needs the total number inspected; a mass fraction needs the total assessed mass and the relevant fraction mass. If the purchase specification uses mass, a count-based result cannot be quietly substituted. Record the raw observations needed to understand the calculation, including any agreed exclusions or preparation steps.

Photographs should make the finding easier to verify. Include the sample identification, an overall view and a close view of the feature being reported. Use a suitable measurement reference where dimensions matter. Keep enough context in the close view to identify the sample and the part being examined. The picture should support the written observation, not replace the measurement record.

Check packing at the relevant levels: the inner pack, carton and pallet arrangement where included in scope. Compare the actual coding, product description and marks with the approved reference. Record the nature and location of a damaged closure or unclear code. A generic note that “packing is good” cannot tell the order team whether the required label version or lot code was actually checked.

If a prepared sensory check is required, use the agreed preparation and food-handling procedure. State what was assessed and when. Do not infer a flavour or texture result from the frozen photograph, and do not let an acceptable sensory result stand in for a laboratory test that was separately required.

Keep temperature checks connected to their locations

Record what each temperature figure describes. Cold-room air, a reading between packs and a measurement in contact with product have different targets. A photograph of a display without its location, instrument identity and time leaves the buyer unable to interpret it. The specification and inspection brief should establish which evidence is required.

The Codex code for processing and handling quick-frozen foods describes a stepwise inspection approach: examine visible condition and temperature records, then use appropriate product measurements when necessary. It also calls for suitable equipment and calibration or checks. These principles support a reasoned assessment; they do not make a single air reading proof of every carton’s condition.

For each measurement, keep the sample or stock location, target, method, units, time and instrument reference together. If a required location cannot be assessed, say so. If the reading prompts further examination, link the follow-up to the original observation. Record a between-pack reading as a between-pack reading rather than relabelling it as a core measurement.

Arrange the work to protect frozen product condition and minimise unnecessary exposure during inspection. The responsible facility and inspector should agree the working area and handling sequence. For any concern, retain the affected identity and relevant history for assessment; cooling the material again does not explain what happened before the reading.

Use the agreed product and destination requirements when assessing a deviation. Check that any cited temperature tolerance applies to the assessed stage, product and jurisdiction. If the evidence is incomplete, identify the missing record or measurement and the person responsible for deciding what happens next.

Protect samples intended for laboratory testing

Arrange the laboratory work early enough for its outcome to serve the release decision. Confirm the test, method, material, collection plan, required sample quantity and expected reporting date with the laboratory. The product’s intended use and the relevant hazard or compositional question determine the required assessment; the test scope therefore needs to match the actual product and order.

FDA’s Bacteriological Analytical Manual, Chapter 1 explains why representative collection and sample condition matter to microbiological results. It addresses protected collection, identification, maintaining frozen samples during transport and documenting receipt condition. Follow the laboratory’s suitable collection and handling instructions; a sample previously opened and handled for a visual display should not be assumed suitable for microbiological testing.

Prepare a handoff record that connects the sample code to the product, lot, collection date, collector and requested analysis. Record who receives the sample and any condition problem reported at receipt. If the laboratory cannot analyse it because the container or sample condition is unsuitable, record that outcome as a sampling failure requiring resolution, not a passing test.

If final laboratory results are a condition of release, keep the affected lot on hold until those results are available and reviewed by the responsible person. Keep requested, received, in progress and final-result status distinct. A booking confirmation proves that work was arranged; it does not contain the result. A report must identify the submitted material and the tests actually performed. Review it against the agreed scope and retain any interpretation or further action needed from the responsible QA person.

Do not use repeated testing merely to replace an unwanted result with a favourable one. In its guidance on testing for Salmonella, FDA recommends maintaining control of tested food pending the final outcome. It also explains that a valid confirmed positive is not erased by later negative tests unless circumstances demonstrate the first result was inaccurate. That is specific FDA guidance; the applicable investigation and disposition must follow the product’s requirements and the responsible food safety decision.

Build a report that supports a release decision

The report should let someone who was not present reconstruct the inspection. Identify the order, goods presented, lot breakdown, inspection date, location, specification revision and inspection party. Show the checks completed, their methods and results, and the evidence supporting each significant finding. State omissions and pending work clearly.

A practical report line places the requirement beside the finding. If an agreed dimension was checked, identify the observed values and method. If a label version was wrong, identify the version found and the stock affected. If a test is pending, leave its status pending. Keep the supporting observations beside any summary tick or status.

Agree who can make the commercial and QA decisions before the visit. The inspector’s findings and the buyer’s release authority may be different responsibilities. A report issuer should not be assumed to have authority to waive an order requirement, and a shipping deadline should not silently change a required acceptance condition.

For a nonconformity, record the affected material, corrective action proposed and verification needed. A replaced label, reworked lot or substituted product requires evidence that addresses the actual finding. Preserve the original report and connect the follow-up to it so the reason for the final decision remains understandable.

Carry the accepted lot into the loading handoff

Pre-shipment inspection and loading are separate moments. The later handoff must still establish which cartons or pallets were loaded and how they relate to the released goods. If the stock allocation changes after inspection, review the change before treating the earlier report as applicable to the replacement material.

Match the accepted lot references and quantities to the loading record. Keep the container identification, seal reference, date and relevant shipment documents together. Where container condition, set point or other loading checks are required, define who records them and what evidence is retained. Our cold-chain and container-loading coordination connects these requirements with the responsible facility and logistics parties.

For a partial release, identify the permitted material precisely. A statement that “the order passed” can cause confusion if only some lots were accepted or a replacement remains unreviewed. The warehouse and loading team need an instruction they can apply to the stock in front of them, with a clear route for resolving mismatches.

Retain the final inspection, follow-up and loading references in one order file. For a receiving query, use those references to locate the inspected material, the findings and the decision taken before dispatch.

References

Codex Alimentarius, General Guidelines on Sampling, CXG 50-2004, revised 2023; Codex Code of Practice for the Processing and Handling of Quick Frozen Foods, CXC 8-1976, revised 2008; FDA Bacteriological Analytical Manual, Chapter 1, April 2022; and FDA’s March 2012 Salmonella testing guidance support the specified sampling, handling and interpretation distinctions. SGS’s agricultural pre-shipment inspection service scope provides a commercial example of separately defined quantity, sampling, packing, documentation and testing work. Examples and illustrations here are explanatory, not actual inspection results or prescribed sampling plans.