Information date: 3 September 2026 — China divides medical devices into Classes I, II and III by risk. A foreign manufacturer should not start from a competitor’s product card; it should define intended purpose, structure, contact, duration, energy and the consequence of failure, then test those facts against the official rules and catalogue.
Verified facts and scope
The NMPA classification rules state that Class I, II and III represent increasing risk. Risk is assessed comprehensively using intended purpose, structural characteristics, use pattern, use status and body contact.
Where more than one class appears applicable, the highest risk class generally governs. The rules also address kits, accessories, active devices, absorbable products, measuring functions and devices intended to treat disease.
How the effect reaches operations
Classification changes registration or filing route, evidence, testing, local responsibility and launch time.
A translated brochure that broadens intended use can move the risk analysis away from the home-market dossier.
Treating software, accessories or sterile supply as an afterthought can invalidate the initial plan.
Decision
Freeze a China intended-purpose statement before budgeting. Use the catalogue and rules together, and request formal clarification when a genuinely new product is not listed. Do not promise a launch date while classification, mandatory standards or clinical-evidence route remains unresolved.
Implementation checklist
- Write the intended user, patient, condition, function, environment and claimed outcome in one controlled statement.
- Describe activity, invasiveness, contact site, duration, sterility, measurement and consequence of malfunction.
- Search the current classification catalogue and record candidate codes and competing rationales.
- Check kits, accessories, software and drug-device combination features separately.
- Map the likely class to filing or registration, testing, evidence and local-agent tasks.
- Obtain a second regulatory review before final labels, contracts or launch dates.
Evidence and review
For “China medical-device classification guide: Intended purpose comes before the product brochure”, start with one real case rather than an abstract checklist. Record the input version, responsible owner, start time, observed result and stop condition. If the team cannot complete “Write the intended user, patient, condition, function, environment and claimed outcome in one controlled statement.” with current evidence, it should not expand the process to more products, patients, suppliers or markets. The first review should focus only on facts capable of changing the decision.
The second control follows “Describe activity, invasiveness, contact site, duration, sterility, measurement and consequence of malfunction.”. Keep the source date, applicable population or entity, deadline, cost effect and owner in the same evidence file. A wording preference does not justify a new version. A repeated discrepancy, an unsupported health claim or a regulatory mismatch does: correct that point and hold release until the evidence is available.
After “Search the current classification catalogue and record candidate codes and competing rationales.”, compare the intended outcome with what actually happened. Apply the same success criteria to each later expansion. If only one number, date or responsibility changes, update that field and the affected conclusion instead of recreating evidence that remains valid. This keeps the decision traceable without turning review into an open-ended rewrite cycle.
Release criterion
The release test for “China medical-device classification guide: Intended purpose comes before the product brochure” is not document volume. Each material number needs a date and denominator, each action needs an owner and trigger, and each exception needs an escalation route. When the source, operating step and limit align, minor wording differences do not justify another rewrite. If the conclusion still depends on an unverified assumption, narrow the claim or pause the affected decision until direct evidence is available.
Counter-scenario and ownership
The review must also test the opposite of the expected outcome. If “Classification changes registration or filing route, evidence, testing, local responsibility and launch time.”, the record should already identify who detects it, who can pause the process, and who communicates with affected people or authorities. Direct, current evidence about the studied product, population or transaction takes priority when it conflicts with a broad market statement. Keep both the approval reason and the rejection reason. Later evidence should reopen only the affected question, not trigger an unsupported rewrite of findings that still hold.
Limits of the conclusion
The English NMPA page notes that the Chinese text prevails where interpretation differs. Classification can change with claims and design, and this article does not replace an NMPA confirmation or product-specific regulatory opinion.
