US FDA device classification guide for medical device teams

What US FDA device classification decides

US FDA device classification is the risk-based system used to determine the level of regulatory control needed to provide reasonable assurance of a medical device’s safety and effectiveness. Under section 513 of the Federal Food, Drug, and Cosmetic Act, devices are assigned to Class I, Class II, or Class III.

In practical terms, classification affects whether a product may be 510(k) exempt, requires a 510(k), may qualify for De Novo classification, or requires premarket approval. It also influences labeling expectations, quality system obligations, special controls, clinical evidence needs, and postmarket duties. For regulatory teams, classification should not be treated as a paperwork step after design is complete. It is an early decision that can affect development timelines, evidence planning, and market-entry strategy.

For more regulatory context and related updates, visit the site’s Regulatory section.

The three FDA device classes and what they mean

FDA classifies medical devices according to the level of control needed for safety and effectiveness. The decision is not based only on product complexity, price, or whether the device uses software. A simple-looking device can raise significant safety questions if its intended use is critical. At the same time, a technically advanced product may fit an established device type with defined controls.

FDA class	General risk concept	Main controls	Typical premarket position
Class I	Low to moderate risk	General controls	Many are exempt from 510(k), but not from all FDA requirements
Class II	Moderate to high risk	General controls and special controls	Most require 510(k), though some are 510(k) exempt
Class III	High risk	General controls and premarket approval	Most require PMA unless a specific exception or historical status applies

General controls are baseline requirements that may include establishment registration, device listing, labeling controls, prohibitions against adulteration and misbranding, records and reports, and quality system requirements unless an applicable regulation provides an exemption. Special controls are additional requirements used mainly for Class II devices. They may include performance standards, special labeling, postmarket surveillance, patient registries, guidance recommendations, or specific testing expectations.

Class III is reserved for devices where general and special controls are not sufficient to provide reasonable assurance of safety and effectiveness, or where there is insufficient information to make that assurance. Many Class III devices are life-supporting, life-sustaining, implanted, or present a potential unreasonable risk of illness or injury. For these products, a PMA generally requires valid scientific evidence rather than a substantial-equivalence argument alone.

How to determine the likely classification

FDA’s classification approach starts with the device’s intended use and technological characteristics. A team should describe what the product is intended to do, who will use it, the patient population, the clinical setting, the mode of action, whether it is implanted or life-sustaining, and whether failure could cause serious harm. This definition matters because the same general technology can fall into different classifications when the intended use changes.

After the intended use is clear, the usual next step is to search FDA classification resources. FDA has established classifications for approximately 1,700 generic device types, organized across 16 medical specialty panels in the Code of Federal Regulations. Most medical device classifications are found in 21 CFR Parts 862 through 892. The FDA Product Classification Database connects device names, product codes, regulation numbers, medical specialty panels, review organizations, and other regulatory information.

A practical classification review should document the search terms used, matching product codes, the regulation number, the device class, whether the product is 510(k) exempt, and any limitations on exemption. It should also compare the proposed device against the regulation’s description, not only against a device name that appears similar. Product code names are useful, but the regulation text, intended use, and special controls are what give the classification analysis substance.

A working classification file should include

• The proposed intended use and indications for use.
• A concise technology description, including software, energy source, materials, accessories, and patient-contacting components where relevant.
• Potential matching regulation numbers and product codes.
• Known predicate devices if a 510(k) route appears possible.
• Applicable special controls, recognized standards, and guidance documents.
• Any reason the device may exceed the limits of a 510(k) exemption.
• Open questions for FDA interaction, such as a 513(g) request or pre-submission discussion.

Classification and the premarket pathway are related but not identical

Classification strongly influences the premarket pathway, but it does not answer every regulatory question by itself. A Class I device may still need to comply with registration, listing, labeling, quality, and reporting requirements. A Class II device may be 510(k) exempt only if it stays within the boundaries of the exemption and the classification regulation. A Class III device usually points toward PMA, but historical preamendments status and specific FDA actions can affect the exact path.

Situation	Common pathway implication	Key caution
Class I and listed as 510(k) exempt	No 510(k) submission is usually required before marketing	Other general controls still apply unless specifically exempted
Class II with predicate device	510(k) is commonly used to show substantial equivalence	Special controls and performance data may be central to the submission
Novel low- to moderate-risk device with no predicate	De Novo may be appropriate	The request asks FDA to classify the device into Class I or II based on risk
High-risk device without sufficient general and special controls	PMA is generally expected	Clinical and nonclinical evidence requirements are usually more demanding

The De Novo pathway is important for innovative devices that do not have a legally marketed predicate but may not require Class III PMA-level controls. If granted, a De Novo request creates a new classification regulation and may create a predicate for future 510(k) submissions when applicable. This is different from a 510(k), which depends on substantial equivalence to an existing legally marketed device.

If a team is uncertain whether a product is a device, whether an existing classification applies, or which requirements may apply, a 513(g) request can be used to ask FDA for information. However, FDA guidance states that a 513(g) response is not a marketing clearance or approval. It should be treated as regulatory information, not as authorization to commercialize.

Why 510(k) exemption should be checked carefully

One common classification mistake is assuming that Class I automatically means no FDA submission and no continuing obligations. FDA explains that most Class I and some Class II devices are exempt from premarket notification requirements, but exemptions are limited. A device that is exempt from 510(k) must still comply with other applicable requirements unless a regulation expressly provides an exemption.

Exemption limits matter because they can be triggered by changes in intended use, technology, operating principle, or risk profile. For example, a device may fit an exempt generic type only when it has characteristics that are already commercially distributed and reasonably foreseeable for that type. If the product uses a new technology that raises different questions of safety or effectiveness, a 510(k) may still be needed even if the generic category is often exempt.

For Class II devices, special controls often define the evidence boundary. They may tell the manufacturer which performance characteristics, labeling elements, or risk mitigations FDA expects. Ignoring special controls can lead to a weak submission strategy because substantial equivalence is not only about finding a predicate. The submission must also address whether the new device is as safe and effective as the predicate and whether any technological differences raise different questions of safety or effectiveness. See also: Implants.

Current quality system context after February 2, 2026

Classification also intersects with quality system planning. Since February 2, 2026, FDA’s Quality Management System Regulation has been effective for medical device current good manufacturing practice requirements under 21 CFR Part 820. The QMSR incorporates ISO 13485:2016 by reference and replaced the older Quality System Regulation structure. FDA also stopped using the Quality System Inspection Technique and moved to an updated inspection process under its medical device compliance program.

This quality system change does not replace device classification. A Class I, Class II, or Class III determination still drives the applicable regulatory controls and premarket expectations. However, the QMSR context matters because development teams often need to plan classification, design controls, risk management, supplier controls, production controls, complaint handling, and postmarket processes together. A device that appears simple from a market perspective can still require disciplined quality documentation if it is a finished device intended for commercial distribution.

The practical point is straightforward: classification should be connected to product lifecycle planning. The earlier a company understands its likely class, special controls, submission pathway, and quality obligations, the easier it is to avoid late-stage redesign, missing test data, or unsupported labeling claims.

Common classification errors that create regulatory risk

Several errors appear repeatedly in device classification work. The first is relying on a competitor’s product code without comparing the exact intended use and technology. Product codes are helpful, but they are not interchangeable labels that can be copied across products. A difference in indication, anatomical site, user population, algorithm function, patient-contact duration, or level of automation can change the analysis.

The second error is treating software as a single category. Software can support administrative functions, drive clinical decisions, control hardware, analyze physiological signals, or provide diagnostic information. Classification depends on what the software does and on the risk if it fails or produces incorrect information. The presence of software alone does not answer the classification question.

The third error is overlooking accessories. An accessory may have a different risk profile from the parent device, and FDA has mechanisms to classify certain accessories distinctly when appropriate. Teams should evaluate whether an accessory has its own intended use, risk controls, and regulatory classification rather than assuming it automatically inherits every aspect of the parent device’s classification.

The fourth error is assuming that classification is fixed forever. FDA can reclassify device types when new information, experience, or regulatory controls justify a change. Manufacturers should monitor relevant FDA actions, classification regulations, guidance updates, recognized standards, and product-specific safety signals throughout the device lifecycle.

Practical checklist before choosing a pathway

1. Confirm that the product meets the legal definition of a medical device and is not primarily regulated as a drug, biologic, or another product type.
2. Write the intended use and indications before searching for a product code.
3. Search FDA classification resources by device name, generic technology, product code, panel, and regulation number.
4. Read the applicable 21 CFR classification regulation, not only the database summary.
5. Check the class, product code, review panel, 510(k) exemption status, and exemption limitations.
6. Identify special controls, relevant standards, guidance documents, and known predicates.
7. Assess whether technological differences raise new safety or effectiveness questions.
8. Consider De Novo when the device is novel, lacks a predicate, and appears suitable for Class I or II controls.
9. Consider PMA planning early when the device is high risk or when general and special controls appear insufficient.
10. Use FDA interaction tools when classification remains uncertain, while remembering that information requests are not marketing authorization.

Frequently asked questions

Does Class I mean a device is unregulated?

No. Class I devices are generally subject to general controls unless a specific exemption applies. Many Class I devices are 510(k) exempt, but manufacturers may still need to meet requirements such as establishment registration, device listing, labeling rules, quality system requirements, and records or reporting obligations.

Is a 510(k) always required for Class II devices?

No. Most Class II devices require a 510(k), but some are exempt. The answer depends on the classification regulation, product code, exemption status, and whether the proposed device stays within the limits of the exemption.

When is De Novo classification useful?

De Novo may be useful for a novel device that has no legally marketed predicate but can be regulated through general controls alone or through general and special controls. If granted, it classifies the device into Class I or II and may create a future predicate for similar devices.

Can FDA classification change after a device type is already on the market?

Yes. FDA has reclassification processes and may change the regulatory class of a device type when the available evidence and regulatory controls support a different level of oversight. Manufacturers should monitor updates that affect their product codes and device categories.

What is the most important first step in classification?

The most important first step is defining intended use clearly. Device classification depends heavily on what the product is intended to do, who uses it, where it is used, and what risks arise if it fails or performs incorrectly.