The FDA uses two words that sound interchangeable but are not: a device can be cleared or it can be approved. The difference is not cosmetic. It reflects which of two premarket pathways the device went through, how much evidence the manufacturer had to bring, and what question the FDA actually answered. Understanding 510(k) versus PMA is the cleanest way to read a device's regulatory pedigree.

Start with the 510(k), the premarket notification pathway. Its question is comparative: is this new device substantially equivalent to a device already legally on the market, called a predicate? If the new device shares the predicate's intended use and either the same technological characteristics or different ones that raise no new safety-and-effectiveness questions, the FDA can find it substantially equivalent and issue a clearance. The submission requirements live in 21 CFR Part 807, and the substantial-equivalence comparison, identifying the predicate and showing the match, is the core of the package. This pathway carries most Class II devices, the large middle tier of moderate-risk products.

The PMA, premarket approval, asks a fundamentally different question: does the evidence show that this device is safe and effective, on its own merits? There is no predicate to lean on. The scope of the pathway is set in 21 CFR 814.1, which says the part implements sections 515 and 515A of the act and applies to Class III devices not otherwise exempt. Class III is the highest-risk category, the devices that are life-supporting, life-sustaining, or of substantial importance in preventing impairment of health, where general and special controls are not enough to assure safety.

"This part applies to any class III medical device, unless exempt under section 520(g) of the act, that: (1) Was not on the market ... before May 28, 1976, and is not substantially equivalent to a device on the market before May 28, 1976."— 21 CFR 814.1, source

What the evidence burden looks like in each pathway

The practical gap between the two is the evidence. A 510(k) is built around a comparison; in many cases it can be supported by bench testing, performance data, and the predicate's existing record, without a new pivotal clinical trial. A PMA, by contrast, typically requires valid scientific evidence, often including clinical investigations conducted under an investigational device exemption, that independently establishes the device's safety and effectiveness for its intended use. That is why a PMA review is more involved and why the FDA reserves the word 'approval' for it. The agency is making a stand-alone scientific judgment, not a comparative one.

The classification system is what routes a device to one pathway or the other. Under section 513 of the act, every device type is assigned to Class I, II, or III based on the controls needed to assure its safety and effectiveness. Class I devices are subject to general controls and are frequently 510(k)-exempt. Class II devices are subject to special controls and usually require a 510(k). Class III devices generally require a PMA. So the first question for any device is not 'which form do I file' but 'what class is this device,' a determination anchored in the FDA's classification database and the regulation numbers it assigns.

Reading 'cleared' versus 'approved' correctly

Because the two pathways answer different questions, the labels carry different information. 'FDA cleared' tells you the device went through 510(k) and was found substantially equivalent to a predicate. 'FDA approved' tells you the device went through PMA and the FDA made an affirmative determination, on the device's own evidence, that it is safe and effective. Neither label means a device is risk-free, and neither ranks a device against competitors. They describe the regulatory standard the device met, not a quality grade.

There is also a middle path worth naming, because it sits between these two. When a device is genuinely novel, so there is no predicate to support a 510(k), but its risk is low enough to fit Class I or II, the manufacturer can use the De Novo classification request rather than jumping to a PMA. De Novo creates a new classification and, often, a new predicate that later devices can cite. But the two primary, high-volume pathways remain the 510(k) and the PMA, and the cleanest summary is the one the FDA's own vocabulary encodes: equivalence and clearance on one side, independent evidence and approval on the other.

For anyone evaluating a device, the regulatory takeaway is concrete. Check the device class first, because it predicts the pathway. If the device is Class II, expect a 510(k) clearance grounded in a predicate. If it is Class III and post-1976 without an equivalent, expect a PMA grounded in the device's own safety-and-effectiveness evidence under 21 CFR Part 814. The words on the label, cleared or approved, are not marketing; they are the FDA telling you which question it answered.

Why the 1976 date keeps appearing

Both pathways trace to a single moment: the Medical Device Amendments of 1976, which gave the FDA premarket authority over devices. The scope language in 21 CFR 814.1 is built around that date, applying the PMA pathway to Class III devices that were not on the market before May 28, 1976, and are not substantially equivalent to a pre-1976 device or to one later classified into Class I or II. The 510(k) system rests on the same baseline from the other direction: substantial equivalence is, ultimately, equivalence traced back to a device legally marketed before that date or to the chain of clearances that descends from it. So the two pathways are not unrelated silos. They are two halves of one framework, one for devices that can lean on the existing universe of marketed devices, one for those that cannot and must stand on their own evidence.

That shared origin also explains a practical asymmetry readers notice in the public data. The openFDA 510(k) dataset holds more than 173,000 substantially-equivalent decisions, an enormous volume, because the comparative pathway carries the bulk of moderate-risk devices and recycles predicates efficiently. PMAs are far fewer, because Class III devices are fewer and each PMA is a heavier, evidence-intensive review. The numbers themselves tell the story of the two pathways: a high-volume equivalence lane and a smaller, more demanding original-evidence lane, sorted by the device class that 21 CFR 860.3 assigns at the very start.