Medical Device Clinical Trials in Korea — When They Are Needed and What to Settle First

Key takeaway — A clinical trial is not a step every product goes through. What actually decides it is not the class but whether existing evidence can explain safety and performance. Where one is needed, Article 10 of the Medical Devices Act puts plan approval by the Minister of Food and Drug Safety in front of it — and the same applies to amendments. This is also the largest variable in the schedule. Current as of September 2026.

First — "do we need a trial?"

Class does not answer this. Class III does not always mean yes, and Class II does not always mean no.

The substantive question is: can the evidence you hold explain safety and performance?

The material for that explanation is generally threefold:

Material	What it does
Comparison with an approved product	Shows substantial sameness in principle, structure and intended use
Literature	Assembles published evidence for the technology and indication
Non-clinical data	Explains through laboratory evidence — performance testing, biological safety and so on

Where those three complete the explanation, no trial is needed; where a gap remains, clinical evidence is what fills it. Higher classes make the explanation harder and so raise the likelihood — the class does not create the obligation directly.

So the first task in practice is not requesting a trial quotation but identifying which part of your product existing evidence does not explain. For a genuinely new product that gap is large; for an improvement on an existing product it may sit only in what was improved.

If needed — plan approval comes first

A trial does not simply begin.

A person who intends to conduct a clinical trial with a medical device shall prepare a clinical trial plan and obtain approval from the Minister of Food and Drug Safety; the same shall apply when the plan is amended. Provided that this shall not apply to clinical trials prescribed by Ordinance of the Prime Minister, such as observing the clinical effect of a marketed medical device within its approved particulars, or cases posing little risk of harm to trial subjects. — Medical Devices Act, Article 10 (Approval of Clinical Trial Plans, etc.)

Three parts of that matter in practice.

1. Approval precedes everything. You cannot start by designing the study and engaging sites. Plan approval sits in front, and its duration enters the schedule first.

2. Amendments are also approved. Studies in progress often need the plan adjusted, and each adjustment carries the approval step again. Build that into the plan.

3. There are exceptions. Observing the clinical effect of a marketed device within its approved particulars, and cases posing little risk to subjects, among others prescribed by Ordinance of the Prime Minister, fall outside the approval requirement. But "our study is exempt" has to be confirmed against the ordinance's requirements, not assumed.

Institutions and management standards are also set in law

A trial is not simply a matter of finding a willing hospital.

Article 10(7) provides that matters including what the plan must contain, the content, timing and method of subject consent, the standards for conducting trials, the designation criteria and procedure for trial institutions, and the trial management standards are to be prescribed by Ordinance of the Prime Minister.

Two practical implications:

• Institutional requirements exist. Confirm that a prospective partner site meets them before committing
• Management standards exist. How the trial is conducted and recorded is prescribed, so a study designed without knowing them may not be usable as evidence later

Effect on the schedule — the biggest variable here

A product that needs a trial and one that does not have business plans of different sizes.

The reason is that each stage sits on another party's calendar:

1. Plan approval — the regulator's calendar
2. Site engagement and institutional review — the hospital's calendar
3. Subject recruitment — the least predictable stretch
4. Conduct
5. Reporting and packaging the data

Step 3 especially. The narrower the target condition and criteria, the longer recruitment runs, and this stretch is hard to shorten with budget.

So decide early. Finishing the technical file and only then learning that clinical evidence is required means those five stages start from that point.

For in vitro diagnostics the name differs

In the same position, in vitro diagnostics use a different name and a different approval structure: the clinical performance study. Different terminology means different regulations, so for those products start with the in vitro diagnostics guide.

Relationship to the other files

A trial does not substitute for other evidence. Weak non-clinical data cannot be repaired with clinical data; the reverse is what holds — the denser the non-clinical evidence, the lighter the clinical burden.

• Building laboratory evidence for performance and safety: writing the technical file
• How far the manufacturer's overseas data can be used: foreign test reports guide
• Recording that risk was controlled by design: risk management guide

Common mistakes

• Concluding whether a trial is needed from the class alone
• Requesting trial quotations before identifying what existing evidence fails to explain
• Leaving plan approval time out of the schedule
• Not accounting for approval following plan amendments
• Designing a study without checking institutional and management requirements, leaving the data unusable
• Applying general device trial rules to an in vitro diagnostic product

Before you start

• Identify what existing evidence (equivalence, literature, non-clinical) does not explain
• Decide whether that gap is closed by a trial or by further non-clinical data
• If a trial is needed, put plan approval time into the master schedule first
• Confirm against the ordinance whether the approval exception applies (do not assume)
• Confirm that partner institutions meet the designation criteria
• Design the study against the management standards, so the data is usable
• Allow schedule margin for re-approval on amendment
• For in vitro diagnostics, switch to the clinical performance study rules

---

With clinical trials, the earlier the need is decided, the lower the cost. Send the intended use and a list of the evidence you already hold, and we will identify what remains unexplained and how best to close it — see free preliminary review.