ISO/IEC 17025 clause 7.3: plan sampling and preserve its evidence
Clause 7.3 applies when the laboratory performs sampling for subsequent testing or calibration. It requires a sampling plan and method that control factors affecting the later results. A technically sound test cannot correct an unrepresentative or poorly controlled sampling decision after the event.
What clause 7.3 requires in practice
Clause 7.3 applies when the laboratory itself takes samples that will later be tested or calibrated. It asks for a sampling plan and a sampling method, which together must address the factors that need control for later results to be valid. Both must be available where sampling takes place, and plans should rest on suitable statistical thinking whenever that is reasonable.
The method describes how samples or sites are chosen, the plan, and how the sample is prepared or treated to produce the item for testing or calibration. Records of the sampling data that form part of the work must be kept. Where relevant they cover the method used, date and time, identification and description of the sample, the people and equipment involved, environmental or transport conditions, location sketches where helpful, and any departure from the method or plan. Handling after receipt is then governed by the item-handling clause.
Establish the sampling purpose and method
Clarify what the customer needs the result to represent. Define how samples or sites are selected, the sampling plan and the preparation or treatment needed to produce the item for later testing or calibration. Where reasonable, use appropriate statistical methods in the plan.
The standard does not prescribe a universal number of samples for every material or project. Use the applicable technical method, specification and competent assessment. Make sure the customer and laboratory understand the limits of the conclusion the sampling approach can support.
Make the plan usable where sampling occurs
The plan and method must be available at the sampling site. Personnel need the current instructions, suitable equipment and a way to record the relevant observations. A plan stored only at head office is insufficient if field staff cannot use it when making selection or preparation decisions.
Control relevant conditions and deviations. Consider how sampling connects with item handling under clause 7.4, including transport and preservation. A change in sampling location or preparation can affect the result's meaning and needs appropriate technical evaluation and records.
Retain records that connect the sample to its origin
Clause 7.3.3 requires sampling records relevant to the testing or calibration. These include the method, date and time, sample identification and description, personnel, equipment, relevant environmental or transport conditions, location information where appropriate, and departures from the method or plan.
Use diagrams, coordinates, photographs or another suitable means where location detail is necessary. The format is a practical choice; the record must allow the sampling event and its significance to be understood. Avoid recording only a sample number that cannot be connected to the actual selection and conditions.
Why clause 7.3 matters to the laboratory
A laboratory result describes the sample it received. If the sample was chosen carelessly, no amount of careful analysis can make the answer represent the lot, the stockpile or the site the customer cares about. Sampling is also where evidence is most easily lost, because it often happens away from the laboratory, under time pressure, with only a notebook. Customers building structures, releasing production lots or making compliance claims depend on the link between the sample and the wider material. A well-planned and well-recorded sampling activity lets the laboratory say what its conclusion covers and what it does not, which is the best protection against later disagreement.
How to implement clause 7.3
Design sampling around the conclusion the customer will draw from the results. Field staff are the audience for these controls, so test them in the field, with the equipment and time pressure sampling actually involves.
- Ask during request review what the result is meant to represent: a lot, a location, a period or a single item.
- Write the sampling method for each recurring service, covering selection, preparation and the factors that must be controlled.
- Use a recognised statistical approach for plan design where one fits, and note when practical limits led to a different choice.
- Put the current plan and method into field kits or mobile access so they are in hand at the point of sampling.
- Provide a field record that captures the items the clause lists, with space for location sketches and departures from the plan.
- Train and authorise the people who sample, and check their practice by occasional observation.
- Link each sampling record to the sample identity so the chain to the test report is unbroken.
Two illustrative sampling situations
A construction laboratory samples material from a project site. It records the selected locations, relevant conditions and any departure from the agreed plan. The resulting report does not imply that a convenient sample proves the condition of the entire project without a sampling basis supporting that conclusion.
A manufacturing laboratory samples a production lot for later testing. It follows a defined selection approach and records how the laboratory item was prepared. If production provides a preselected sample instead, the laboratory clarifies that changed responsibility and the limits on what the result represents.
Evidence and common mistakes
The plan, method and relevant sampling records are explicit requirements. A particular branded form or digital application is not. Link the records with sample identification, technical work and reporting so an assessor or reviewer can follow the chain.
Avoid treating sampling as transport administration, changing selection to suit convenience without evaluation, or confusing customer-provided sample information with laboratory observations. Where the laboratory is responsible for sampling, the reporting requirements under clause 7.8.5 also need attention when necessary for interpretation.
Frequently asked questions
Does clause 7.3 apply when customers supply all samples?
The laboratory's sampling duties apply when it performs sampling. It still needs appropriate item handling and accurate reporting of responsibility when customers provide samples.
Must every sampling plan use a statistical method?
The requirement is to use appropriate statistical methods whenever reasonable. The approach must fit the activity, applicable requirements and the conclusion the sampling is intended to support.
Can sampling records be completed later from memory?
Relevant observations should be captured reliably as the work occurs and linked to the task. Reconstructing critical selection or condition information later risks losing the evidence needed to interpret results.
Turn this requirement into reliable laboratory practice
Sterling can help connect sampling responsibilities, field records and reporting limits within your laboratory programme. Bring a representative sampling service to a free consultation to identify the practical controls and training your team needs.
Build the team's understanding through Method Validation and Verification (ISO/IEC 17025:2017). Discuss the relevant scope, employee learning and agreed access to Effivity ISO & GRC automation software with Sterling; laboratory technical decisions remain with competent, authorised personnel.
ISO/IEC 17025 requirements hub · Laboratory process requirements · Protect sample integrity · Report sampling information