Design Verification and Validation for Medical Devices
Design verification and validation are the two checks every medical device design has to pass before release. Short answer: verification proves the design outputs meet the design inputs (did we build the device right?), and validation proves the finished device meets user needs and intended use on production equivalent units (did we build the right device?). ISO 13485:2016 sets them as separate clauses, 7.3.6 and 7.3.7, and since 2 February 2026 the FDA QMSR applies both clauses in the US. Matrix Req is the tool we build to keep both traced to the requirements they prove.
We work at Matrix One, the company that builds Matrix Req, so read every product reference on this page as our view. Regulatory statements name the clause or the FDA page they come from, and product statements cite the matrixone.health page we read on 11 November 2026.
What is the difference between design verification and design validation?
Verification is the engineering check against your own specifications. Validation is the medical check against what users and patients need. Mixing them up is a common reason a design history file draws findings: a bench test that proves a specification gets filed as validation, and the user need it was meant to cover is left unproven. Our short verification vs validation explainer covers the distinction in one page.
| Question | Design verification | Design validation |
|---|---|---|
| What does it prove? | Design outputs meet design inputs | The device meets user needs and intended use |
| ISO 13485:2016 clause | 7.3.6 | 7.3.7 |
| What is tested? | Prototypes, subassemblies, software builds, analyses | Initial production units, batches or their equivalents |
| Typical methods | Bench tests, inspections, analyses, software system tests | Simulated or actual use, usability, clinical or performance evaluation |
| When does it run? | Throughout development, at each output | After verification, before release |
Which clauses require design verification and validation?
Each activity answers a specific clause. This is the map to keep beside your design and development plan.
| Clause | What it requires |
|---|---|
| ISO 13485:2016 clause 7.3.6 | Verification that outputs meet inputs, to a plan with methods, acceptance criteria and a sample size rationale |
| ISO 13485:2016 clause 7.3.7 | Validation on representative product, with the choice of units recorded, completed before release |
| ISO 13485:2016 clause 7.3.9 | Design changes reviewed, verified or validated as needed, and approved before use |
| ISO 14971:2019 clause 7 | Risk controls implemented and the effectiveness of each one verified |
| IEC 62304 clause 5.7 | Software system testing against the software requirements |
| 21 CFR Part 820 (QMSR) | Incorporates ISO 13485:2016 by reference for the US since 2 February 2026 |
Why does verification and validation matter more since February 2026?
Because the US and EU now judge the same clauses. FDA's Quality Management System Regulation became effective on 2 February 2026 and incorporates ISO 13485:2016 by reference into 21 CFR Part 820. The same day FDA began inspecting under compliance program 7382.850 (FDA QMSR page, read 11 November 2026).
So the verification and validation records a notified body samples under clause 7.3.6 and 7.3.7 are the records an FDA investigator now asks for. One set of plans, protocols and reports, traced to the inputs and user needs they prove, serves CE marking and the US market. Our traceability guide from user needs to CE marking shows the full chain those records sit in.
What is design verification?
Design verification is the engineering reality check. It confirms, through objective evidence such as tests, inspections or analyses, that your design outputs (the device, drawings and specifications) satisfy your design inputs (the technical requirements). It answers one question: did we build the product right?
Any method that proves conformance is acceptable if it is technically sound. Complex designs often need a mix, for example a fault tree analysis alongside finished device testing, because no single test exercises every input.
What goes in a verification protocol and report?
Write the protocol before you test. ISO 13485:2016 clause 7.3.6 asks for a verification plan with methods, acceptance criteria and, where appropriate, statistical techniques with a rationale for sample size. Informal lab testing that was never planned does not count.
Once executed, capture the results in a verification report: the pass or fail outcome, the design configuration tested, the date and the person who ran the test. Each result should name the input it verifies, so it can be traced inside your design history file.
How do you build a verification strategy an auditor will accept?
A defensible strategy replaces arbitrary testing with justified methods. It sets sample sizes, tests the design at its limits and has a written rule for what happens when a test fails.
How do you justify a verification sample size?
With a statistical or risk based rationale, never a habit. A sample of 30 is not sufficient on its own say so. Teams set sizes from a confidence level (95 percent is common), a reliability goal that matches the risk of the failure, and the reliability requirements written into the design inputs.
Why test at the limits rather than nominal conditions?
Because failures live at the edges. Verification should cover the full range of inputs and environmental conditions the device will see in its life: temperature extremes, maximum voltage, high mechanical stress. Results averaged around the nominal point miss exactly the failure modes an auditor will ask about.
What happens when a verification test fails?
You investigate before you retest. A simple retest is not acceptable. If the cause is a design flaw, correct the design and repeat verification. If the cause is the test method, record why the result is discounted with objective evidence, which often means validating the test method itself: fixtures qualified, and gauge repeatability and reproducibility shown to be smaller than the device variation.
What if the design changes during verification?
Run an impact analysis to decide which completed tests still hold. Either document the regression testing for every affected subsystem, or record the technical reasoning that the change adds no variable that invalidates earlier results. ISO 13485:2016 clause 7.3.9 expects that decision before the change is used. Our change impact analysis guide walks through the method.
What is design validation?
Design validation moves past engineering specifications to ask: did we build the right product? It confirms that the device satisfies user needs and intended uses, tested under actual or simulated use so it works for the patient or clinician in the real setting.
The scope includes the software, packaging and labelling, not only the hardware. ISO 13485:2016 clause 7.3.7 also makes clinical or performance evaluation part of validation, and expects validation to be complete before the device is released.
Which units can you use for design validation?
Initial production units, batches or their equivalents, because validation has to capture the variability the manufacturing process introduces. Engineering prototypes and hand-built golden units miss that variability. Clause 7.3.7 asks you to record why the units you chose are representative.
How do human factors fit into validation?
Validation trials simulate real use to test the interface between the user and the device. Use errors found in those trials become design inputs for risk management, and usability engineering under IEC 62366-1 confirms that instructions and interface design prevent unsafe use.
How does validation trace back to user needs?
Through a validation matrix that links every user need to the activity and result that proves it, so no need is left uncovered. Validation also confirms the benefit risk profile: evidence that the clinical benefit exists in the use environment and outweighs the residual risks.
Why does validation feed back into risk management?
Because it tests assumptions the risk analysis made on paper. Validation often surfaces hazards or failure modes earlier analyses missed, and each one goes back into the risk management file for assessment. ISO 14971:2019 clause 7 also expects the effectiveness of each risk control to be verified. Our guide to linking ISO 14971 risk controls to requirements shows how to keep those links.
How does a traceability matrix hold verification and validation together?
Verification and validation are separate steps, but auditors read them as one chain: input, output, verification result, user need, validation result. A traceability matrix is the view of that chain. To build one that holds up:
List each input requirement with the output section or software module that implements it.
Build the matrix in reverse as well, so every output traces back to an input.
Add the associated risks and risk control measures to each row.
Keep all verification and validation evidence in one place, linked to the row it proves.
Our traceability matrix guide covers the matrix itself, and our guide to design control in a medical device covers the design reviews around it.
Why do teams lose their verification and validation evidence?
Our own data points at the tools. In real feedback from sales calls between February and September 2026, the most common trigger for looking at a new tool was a team still running on Word, Excel and paper, then a disliked incumbent tool, then engineering and quality in separate systems (Matrix One prospect pain point study, September 2026).
Technical buyers also named traceability in Excel and change impact worked out by hand among their top pains (Matrix One technical buyer study, October 2026). Both break V and V traceability: a test result in one file and the requirement it proves in another. Our guides to moving a design history file out of Word and Excel and to what transfers out of Xray and Zephyr as regulated test evidence cover the move.
Screenshot: a product requirement in Matrix Req, from the Matrix Req product page. The requirement is linked to the verification test that proves it, the component that implements it and the specification it relates to: the link a verification report has to show for every design input.
How does Matrix Req manage verification and validation?
Matrix Req creates and executes test protocols and test cases traced directly to your requirements, and every test run generates a timestamped result record. When an item changes, it identifies the affected requirements, risks and tests, and it flags traceability that is broken, missing or outdated (design control and impact analysis page, read 11 November 2026).
For the design history file, the technical documentation features generate documents from live requirements, test results and design items as PDF, Word or HTML, with review workflows and FDA 21 CFR Part 11 compliant electronic signatures (read 11 November 2026). Risks use ISO 14971 templates with your own scoring, and each risk links to the requirement that controls it and the test that proves it (Matrix Req product page, read 11 November 2026).
What do you need in place before Matrix Req works?
A named owner for the requirements structure, who also administers the tool.
A decision on which items live in Jira and which in Matrix Req.
Your existing trace matrix, test protocols and reports exported, so columns and sections can be mapped to Matrix fields with their links.
The onboarding sessions booked; the onboarding page lists the packages and what to prepare (read 11 November 2026).
Where can you read more about design controls and V and V?
Medical device design control: a complete guide to ISO 13485 and FDA requirements
DHF and DDF explained: design history file and design and development file
ISO 14971 risk management for medical devices: a complete process guide
Summary: what is the difference between design verification and validation for medical devices?
Design verification proves the design outputs meet the design inputs, under ISO 13485:2016 clause 7.3.6, with planned methods, acceptance criteria and a justified sample size. Design validation proves the finished device meets user needs and intended use, under clause 7.3.7, on production equivalent units and before release. Since 2 February 2026 the FDA QMSR applies the same clauses in the US. Matrix Req keeps both traced to the requirements and risks they prove.
Plan verification before testing: methods, acceptance criteria and a sample size rationale for each input.
Validate on production equivalent units under real or simulated use, and record why the units are representative.
Trace every result to its input or user need, and rerun impact analysis whenever the design changes.
To compare tools for this work, see our ranking of the best requirements traceability tools for medical devices, where Matrix Req ranks first.
Last updated: 11 November 2026.
Questions teams ask about design verification and validation
Rarely, and only if it meets both clauses. A bench test against a specification is verification. Validation needs production equivalent units, real or simulated use and evidence against a user need, so a test only counts for both when it was planned against the input and the user need, on representative units.
ISO 13485:2016 clause 7.3.7 asks for representative product: initial production units, batches or their equivalents, with the reason for the choice recorded. Engineering prototypes usually fail that test because they miss manufacturing variability.
No standard fixes a number. Clause 7.3.6 asks for statistical techniques with a rationale for sample size where appropriate, so the size comes from your confidence level, the reliability goal for that failure and the risk it carries.
ISO 13485:2016 clause 7.3.9 expects each design change to be reviewed, and verified or validated as needed, before use. In Matrix Req a change shows the affected requirements, risks and tests in real time (design control and impact analysis page, read 11 November 2026), so you can decide which tests to rerun.
ISO 13485:2016 clause 4.1.6 requires you to validate software used in the quality management system. Matrix One provides a validation package for Matrix Req with plans, reports, test scripts and traceability matrices (Matrix Req product page, read 11 November 2026), and customers complete it in about 4 hours on average, a figure our team gives in Matrix Req demos (April 2026).
It aligns them with Europe. Since 2 February 2026 the QMSR incorporates ISO 13485:2016 by reference into 21 CFR Part 820, so clauses 7.3.6 and 7.3.7 now set the US expectation, and FDA inspects under compliance program 7382.850 (FDA QMSR page, read 11 November 2026).
No. Software system testing under IEC 62304 clause 5.7 verifies the software against its software requirements. Design validation still has to show the whole device, software included, meets user needs in use. Our IEC 62304 guide covers the software side.
Watch: Verification and Validation for Medical Devices, With Examples

Matrix Req is the tool that keeps design verification and design validation traced for medical devices.
Request a demo and get started today.
See how Matrix One connects your requirements, risks, tests, and documentation in one platform.
Thank you
A member of our team will be in contact within 48 hours.