A PMN flag on a TSCA Inventory record is routinely read backward. It looks like a warning that a premanufacture notice is required. It means the opposite. The substance already went through premanufacture review, someone commenced manufacture, and EPA added it to the Inventory as a result. It is an existing chemical, and you do not file a PMN for an existing chemical.
EPA's glossary puts it in five words: PMN "indicates a commenced PMN substance." The word carrying the meaning is commenced, and it is the word that gets skipped.
The flag appears on 11,383 substances in the July 2026 non-confidential Inventory, which is 16.1 percent of it.
How a substance earns the flag
A new chemical substance, meaning one not already on the Inventory, requires a premanufacture notice under TSCA section 5. Section 720.75 gives EPA a 90-day review period from receipt of a complete notice, extendable under section 5(c), and suspendable at the submitter's request if EPA agrees.
Clearing review is not what puts a substance on the Inventory. Manufacturing does. Under 720.102(a), a person who commences manufacture of a new substance for a nonexempt commercial purpose, having previously submitted a section 5(a) notice, must file a notice of commencement. Section 720.102(b)(1) gives them 30 calendar days from the first day of manufacture, on EPA Form 7710-56 through CDX.
The rule is precise about when that clock starts. The date of commencement is the date of completion of non-exempt manufacture of the first amount, whether batch or drum, of the substance identified in the PMN. For importers it is the date the substance clears United States customs.
EPA then lists the substance, and the listing carries the PMN flag permanently. It is a provenance marker. It records how the substance arrived rather than what may be done with it.
The five ways a review can end
What happens during those 90 days explains most of what you see on the record afterward.
Section 720.75 requires EPA to make one of five determinations under TSCA section 5(a)(3) within the review period. The substance presents an unreasonable risk. The available information is insufficient for a reasoned evaluation. In the absence of sufficient information, the substance may present an unreasonable risk. The substance will be produced in substantial quantities and either will enter the environment in substantial quantities or will involve significant human exposure. Or the substance is not likely to present an unreasonable risk.
Only the last of those five clears the way without conditions. The other four lead to action, most often an order under section 5(e) setting terms for how the substance may be made and used. This is a real change from how section 5 worked before 2016, when a notice could clear simply by the review period running out. EPA now has to decide something.
That structure is visible in the flags. A substance whose review ended in the fifth determination tends to carry PMN alone. One whose review ended in any of the first four tends to carry 5E next to it, and often S later on.
What the flag does not restrict
Of the 11,383 public substances carrying PMN, 9,853 carry no TSCA action flag whatsoever. No significant new use rule, no section 5(e) order, no test rule, no section 6 rule. Eighty-seven percent of commenced PMN substances are ordinary existing chemicals that happen to have entered commerce after 1979 rather than before it.
The flag is also no indication of dormancy. 8,484 of the 11,383 are designated active and 2,899 inactive, which is a healthier ratio than the Inventory as a whole.
That said, the flag keeps company worth checking. 1,306 of these substances also carry the S flag for a significant new use rule and 664 carry 5E for a section 5(e) order. The pattern behind those numbers is the normal outcome of a contested review: EPA reaches an order limiting how the substance may be handled, the substance commences and joins the Inventory, and a SNUR then extends the limits to everyone else. The PMN flag tells you where to look for that history. It does not tell you whether it exists.
More of these substances are secret than are public
This is the part of the PMN story that a CAS-number search cannot show you, and it is larger than most people expect.
Alongside the public Inventory file, EPA publishes a second one for the confidential portion. It identifies each substance by PMN number and accession number with a structurally descriptive generic name, and no CAS number at all. The July 2026 release of that file holds 15,967 entries, and 12,503 of them carry the PMN flag.
Set that against the 11,383 in the public file. Of roughly 23,900 commenced PMN substances on the TSCA Inventory, about 52 percent sit on the confidential portion. For those, there is no CAS number to look up, in our data or anyone's.
The two files reconcile exactly against EPA's own published totals, which is worth stating because it establishes the size of the gap rather than estimating it. 70,774 public substances plus 15,967 confidential gives 86,741, the figure EPA publishes. The activity split reconciles too: 36,522 public active plus 6,176 confidential active is 42,698, and 34,252 plus 9,791 is 44,043. Both match EPA's numbers to the unit.
Why so many are confidential, and how the door opens
Section 720.102(c)(2) explains the concentration. A submitter who wants the chemical identity kept off the public Inventory must claim it as confidential and satisfy the certification, substantiation and generic name requirements of 40 CFR part 703. Otherwise EPA lists the specific chemical identity publicly. New chemicals are where commercial secrecy is most valuable, so that is where the claims cluster.
A claim also comes with a condition that matters to everyone else. Under 720.102(e)(1)(ii), a person asserting confidentiality for chemical identity must agree that EPA may disclose, to a person with a bona fide intent to manufacture or import that substance, the fact that it is on the confidential Inventory, for the purposes of section 5(a)(1)(A) notification.
That is the legal basis for the step we keep pointing at. The confidential portion is not a closed door. It is a door that opens to a specific procedure and not to a database search.
The expensive mistake this prevents
Put the pieces together and a practical rule falls out.
If you search the published Inventory for a CAS number, find nothing, and conclude you have a new chemical requiring a premanufacture notice, you may be wrong about half the time that the substance is a commenced PMN substance. It may be sitting on the confidential portion already, listed, with an accession number instead of a CAS number. A PMN is a substantial filing with a substantial EPA fee attached, and a 90-day review in front of it.
The correct sequence is to establish absence through a bona fide intent submission before treating a search miss as a finding. This is the same caution that applies to reading active and inactive designations, and it has teeth here because the cost of getting it wrong is a filing rather than a delay.
Any tool that reports "not on the TSCA Inventory" from a CAS lookup, ours included, is reporting the absence of a public record. On this particular flag, that is a meaningfully weaker statement than it sounds.
How to check a substance
EPA publishes both files, the non-confidential Inventory and the confidential accession listing, roughly every six months. A CAS search covers the first. Nothing covers the second.
One mechanical caution applies to reading either. EPA delivers the flags as a single semicolon-separated field, so a substance with several appears as text such as "PMN; XU" rather than as separate values, and 4,376 of the public PMN substances do carry XU alongside. A tool that stores that string whole cannot count or filter by flag and will not return this substance when you search for PMN. Published counts vary for exactly this reason.
Ambrose holds the July 2026 release, 70,774 substances with their flags split and searchable, and every lookup in our regulatory compliance platform reports which EPA release the answer came from and says plainly what a miss does and does not mean.