Device classification is the regulatory process of assigning a medical device to a risk category that determines its market pathway. Frameworks like the FDA (Class I, II, III) and EU MDR (Class I, IIa, IIb, III) tie each class to specific controls, evidence, and conformity assessment requirements before a device can be sold.
What is device classification?
Device classification is the first regulatory decision a manufacturer makes. It sorts a device into a risk tier based on intended use, invasiveness, duration of body contact, and the control needed to give reasonable assurance of safety and effectiveness. The class then drives most of what follows: submission type, clinical evidence, quality system scope, and post-market duties.
In the United States, the FDA groups devices into Class I, II, and III under 21 CFR Part 860, with device-specific classification regulations listed in 21 CFR Parts 862 through 892. The European Union uses Annex VIII of EU MDR 2017/745, which applies 22 classification rules to place devices in Class I, IIa, IIb, or III. In vitro diagnostics follow separate schemes: IVDR 2017/746 in Europe (Class A to D) and FDA product codes.
Why device classification matters in medical device development
Class decides the regulatory pathway, and the pathway decides cost and timeline. A Class I exempt device in the US may reach the market with registration and listing only, while a Class III device usually needs a Premarket Approval (PMA) supported by clinical trial data. Getting the class wrong early sends teams down the wrong path.
Misclassification is one of the more expensive regulatory mistakes. Claiming a lower class than warranted can trigger rejection, a warning letter, or withdrawal, while claiming a higher class wastes months on evidence that was never required. Either error raises audit exposure, since reviewers check that the rationale matches the intended use and technology.
Classification also shapes the quality system. Design controls, risk files, and post-market surveillance scale with class, so an accurate call sets a realistic program budget.
How device classification works.
Classification is a structured determination. The core inputs are consistent across regions: intended use, indications for use, and technological characteristics.
A typical path looks like this:
- Define the intended use and indications precisely. A wording change, such as adding a therapeutic claim, can raise the class.
- Identify the technology and contact profile. Determine whether the device is invasive or non-invasive, active or passive, its duration of contact, and anatomical location.
- Apply the regional framework. For the FDA, search the product classification database for a matching product code under 21 CFR Parts 862 to 892; if none exists, a De Novo or 513(g) request may apply. For the EU, work through the 22 rules in Annex VIII (non-invasive Rules 1 to 4, invasive Rules 5 to 8, active Rules 9 to 13, special Rules 14 to 22).
- Confirm the class and map it to the regulatory pathway. Determine whether the device requires 510(k), PMA, or De Novo in the US, or self-declaration or Notified Body review in the EU.
- Document the rationale. Record the classification justification in the technical documentation or Design History File (DHF), since reviewers will ask for it.
Software adds nuance. Under EU MDR Rule 11, clinical decision-support and diagnostic software are often classified as Class IIa or higher. Under the FDA, Software as a Medical Device (SaMD) is classified according to its intended function and associated risk.
Common challenges and best practices
The most common failure is treating classification as a late formality. By the time a class is confirmed, the intended use should be stable, since the two are linked. Teams that draft indications loosely often discover a class change after the design freeze.
A second issue is region-by-region drift. A device can be Class II in the US and Class IIb in the EU for the same intended use, so a single global class is a myth. Map each market separately and keep the rationale in one controlled document.
Good practice is straightforward:
- Fix the intended use early.
- Classify against every target market in parallel.
- Record the reasoning with references to the exact rule or product code.
- For borderline cases, such as devices that could fit multiple classes or drug-device combinations, seek a formal opinion. The FDA offers the 513(g) process, while EU manufacturers can consult their Notified Body or competent authority.
How SJML helps with device classification
SJML provides device classification as part of its Compliance-as-a-Service offering. Its QARA team develops regulatory strategy across FDA, EU MDR, IVDR, and country-specific frameworks, mapping a device’s intended use and technology to the right class and pathway in each market. That work connects to registration support (510(k), CE marking, technical files, and DHF), clinical evaluation planning, and QMS build-out under ISO 13485. Because SJML also handles design, engineering, and manufacturing, classification decisions stay aligned with how the device is built and documented.
Frequently asked questions
The manufacturer determines the classification in the first instance, then justifies it to regulators. In the US, the FDA confirms the class through existing classification regulations, a 513(g) request, or a De Novo decision. In the EU, the manufacturer applies the Annex VIII rules and, for most classes above Class I, a Notified Body reviews the result during conformity assessment.
No. The two systems use different labels and rules. The FDA uses Class I, II, and III based on product codes and 21 CFR Parts 862 to 892. The EU MDR uses Class I, IIa, IIb, and III based on the 22 rules in Annex VIII. A device can land in a different tier in each region, so classification must be performed separately for every target market.
Misclassification can stall or void a submission. Claiming too low a class may lead to rejection, a warning letter, or a recall if discovered after launch. Claiming too high a class wastes time and money on evidence that was not required. Because the class sets the regulatory pathway, an error usually surfaces late and forces rework, so it pays to confirm the class before heavy investment.
Software as a Medical Device (SaMD) is classified according to its intended function and the risk it presents. Under EU MDR Rule 11, diagnostic or treatment decision-support software is often classified as Class IIa, IIb, or III, depending on the seriousness of the health situation and the information provided. Under the FDA, SaMD follows the appropriate classification and regulatory pathway based on its function, using either an existing product code or the De Novo pathway.
Related terms
- Regulatory Pathway
- 510(k) Submission
- Premarket Approval (PMA)
- CE Marking
- Conformity Assessment
