Class III and IV Devices in Korea — How Approval Differs from Class II Certification

Key takeaway — After Class I notification and Class II certification, the last box is Class III and IV approval. What changes is not only the name of the pathway. Class II certification asks "does this product meet the prescribed standards"; Class III and IV approval asks "is there evidence that this product is safe and performs as intended." Change the question and both the documents and the schedule drivers change. Current as of September 2026.

What changes as the class rises

The overall structure of how pathways divide by class is covered in the procedure overview; the table below places Class III and IV within it.

Class	Pathway	Central question
I	Notification	Has the product been notified for what it is
II	Certification	Does it meet the prescribed standards (technical file review)
III · IV	Approval	Is there evidence that it is safe and effective

Up to Class II, the standard exists first and the product is checked against it. Preparation focuses on which standards apply and whether testing was performed under their conditions.

At Class III and IV the focus shifts. Meeting the standards is still required, but on top of that you have to establish that the product is safe for people and does what it claims — because an area opens up where the standards do not supply the answer in advance.

The practical difference between III and IV

Both are approval pathways. Depth of evidence rises with risk, but the axis that actually moves schedule and cost is not the class number.

The real fork is this: can existing evidence explain safety and performance?

• Where principle, structure and intended use are substantially the same as an approved product, that comparison explains much of it
• Where the technology is established and supported in the literature, the explanation is likewise manageable
• Where the principle, the indication or the manner of use is new, a gap remains there

Two Class III products, one in the first situations and one in the last, can differ severalfold in preparation time. So the question to ask right after confirming the class is not "how long does Class III take" but "which part of our product is not explained by existing evidence."

The skeleton of the file carries over

If you have prepared a Class II certification, the axes continue.

• Performance and safety testing to the applicable standards — alignment of standard and test conditions
• Biological safety — endpoints derived from the contact categorisation (ISO 10993 guide)
• Risk management — the chain from hazard identification through control to verification (risk management guide)
• Usability — requirements set by users and use environment (usability guide)
• Technical file — binding the results into intended use – principle – specification – testing (writing the technical file)

On top of that sits the evidence of safety and performance. Whether existing material fills it or clinical evidence is required is covered in the clinical trials guide.

Two variables that set the schedule

"How long does it take" has no single answer at Class III and IV because the timeline is driven not by class but by two variables.

1. Whether clinical evidence is needed. Where a trial is involved, plan approval → site engagement → recruitment → conduct → data packaging stack in sequence, and recruitment in particular is hard to shorten with budget. Switch this variable on and the order of magnitude changes.

2. How many deficiency rounds occur. Deficiencies are closer to routine than exception, and the deficiency period is excluded from the statutory processing time. The tighter the internal logic of the file, the fewer rounds; the more the values disagree across documents, the more.

Both variables are reducible at the design stage — the first by identifying early what existing evidence can cover, the second by aligning specification, standards, test conditions and technical file from the beginning.

Business requirements run in parallel

The heavier the item pathway, the easier it is to defer the business requirements — and that is a common mistake.

Manufacturing or import licensing, the quality manager and KGMP do not overlap in requirements with the item pathway, so they can run alongside it. The longer the item review at Class III and IV, the greater the gain from parallelising. The structure is set out in the manufacturing licence guide.

Common mistakes

• Concluding from the class alone that "Class III means a trial is required"
• Conversely, proceeding without clinical evidence without having identified what existing evidence fails to explain, then stalling on deficiencies
• Carrying Class II experience over without establishing safety and performance evidence as its own axis
• Deferring business requirements because the item review is long
• Submitting with values that disagree across specification, test reports and technical file, producing repeated deficiencies

Before you start

• Fix item name and class (obtain formal MFDS classification confirmation where unclear)
• Identify what existing evidence does not explain — principle, indication or manner of use
• Decide how to close that gap — further non-clinical data, or clinical evidence
• If a trial is needed, put plan approval time into the schedule first
• Align standards, test conditions, specification and technical file from the outset
• Start the risk management file early — assembled late, the causality breaks
• Start business requirements (licence, quality manager, KGMP) in parallel with the item pathway

---

At Class III and IV, what moves schedule and cost is not the class number but how much is explained by existing evidence. Send the intended use and a list of what you hold, and we will identify what remains unexplained and the sensible route to close it — see free preliminary review.