How to determine your PAMA status from your own records

Step 1: fix the entity the tests apply to

The tests do not apply to “your company”. They apply to the entity that bills Medicare Part B under its own NPI. 42 CFR 414.502 (eCFR text as of August 20, 2026, read 2026-08-25) opens the definition of applicable laboratory with an entity that:

with one variant: a hospital outreach laboratory that bills under the hospital’s NPI runs the tests on its Medicare revenues attributed to the CMS-1450 type of bill 14x instead of an NPI. If your organization holds several billing NPIs, each one is tested separately. If it holds one NPI covering several CLIA certificates, the NPI is still the unit.

Write down the NPI (or the 14x revenue stream) before touching any number. Most wrong determinations start by running the tests at the company level instead of the NPI level.

Step 2: confirm the entity is a laboratory under the CLIA definition

The definition 414.502 points to is 42 CFR 493.2 (read 2026-08-25):

The same section excludes pure logistics: “Facilities only collecting or preparing specimens (or both) or only serving as a mailing service and not performing testing are not considered laboratories.” A draw station NPI that performs no testing fails here and the analysis stops for that NPI.

Step 3: build the denominator, total Medicare revenues for the window

Pull every Medicare payment received by the NPI for services furnished January 1, 2025 through June 30, 2025. The regulation defines what counts, inside the definition of the majority test at 42 CFR 414.502:

Two reading notes, both from CMS’s own guidance. First, Medicare Advantage money is out: CMS’s PAMA FAQ (A3.4, updated July 14, 2026) says that “Effective January 1, 2019, Medicare Advantage (MA) plan payments under Medicare Part C are no longer considered ‘Medicare Revenues’” for this threshold. Second, the anchor is the date of service, not the date of payment: the definition says “for services furnished during the data collection period.” A payment received in August 2025 for a test furnished in May 2025 belongs in the window. Your 835 remittance data carries both dates; filter on the service date.

Include the beneficiary deductible and coinsurance amounts associated with those services. The definition names them explicitly, and leaving them out shrinks the denominator and can flip the ratio.

Step 4: build the numerator, CLFS plus PFS revenues in the same window

From the same filtered set, total what Medicare paid under the Clinical Laboratory Fee Schedule and under the Physician Fee Schedule. CMS’s CLFS and PAMA reporting resources page (last modified August 3, 2026, read 2026-08-25) gives the method as arithmetic: “Add the Medicare CLFS revenues + the Medicare PFS revenues you got from your own NPI” during the collection period as the numerator, and “the total Medicare revenues you got from your own NPI” as the denominator.

Step 5: run the majority test

Divide. The same CMS page states the pass condition: “If the resulting number is greater than 0.50, or 50%, you meet the” majority of Medicare revenues threshold. The regulation requires strictly more than 50 percent, from the CLFS (subpart G) or the PFS (subpart B) or the combination.

This is the test no outside party can run for you. CMS publishes what it paid each lab NPI annually, but it publishes no NPI’s revenue composition, so the ratio is invisible from public data. A lab whose NPI also bills meaningful non-lab Part A or Part B services can sit under 50 percent with a large CLFS number; a small pure-play lab is usually far over it.

Step 6: run the low expenditure test

The second test, from the same definition in 42 CFR 414.502: the entity

That is $12,500 of CLFS money alone, during the six month collection period, not per year. One carve-out exists for a single laboratory offering an ADLT: the threshold “Does not apply with respect to the ADLTs it offers and furnishes” and “Applies with respect to all the other CDLTs it furnishes.” If you are not the sole source of an advanced diagnostic laboratory test, the carve-out does not concern you.

Step 7: read the result, and write it down

Both tests must pass, on the same window, for the same NPI. One number at or under its bar means the entity was not an applicable laboratory for this cycle, and the regulation is explicit that nothing should have been filed for it: 42 CFR 414.504(g) says “Applicable information may not be reported for an entity that does not meet the definition of an applicable laboratory.”

Both tests passing means the reporting obligation existed, and it sat one level up: the TIN-level entity reports for its component applicable laboratories, and an officer signs. 42 CFR 414.504(d) requires that “the President, CEO, or CFO of a reporting entity, or an individual who has been delegated authority to sign for, and who reports directly to, such an officer, must sign the certification statement”. The reporting period for this cycle ran May 1, 2026 through July 31, 2026 and is closed.

Either way, keep the worksheet: the filtered revenue extract, the two totals, the ratio, the date you ran it, who ran it. The determination is self-made, so the record of how it was made is the only evidence it happened. The next cycle reuses the same procedure on a future window: CMS’s FAQ says the three year cycle resumes with 2029.

What no outside source can do for this procedure

Public data can put a floor under one of the two tests and nothing more. The checker at pamawatch.com/check shows what CMS’s published payment file says about a given NPI next to the $12,500 bar, and stops exactly where the public record stops: it cannot see the denominator, so it never states a status. The numbers above, out of your own remittance data, are the only place the answer exists.

The paper behind each step

Every text quoted in these steps was fetched on 2026-08-25: the eCFR’s August 20, 2026 versions of 42 CFR 414.502, 414.504, and 493.2; Section 6226 of Public Law 119-75 (140 Stat. 684, enacted February 3, 2026) from govinfo; and CMS’s two operational documents, the reporting resources page in its August 3, 2026 revision and the FAQ document in its July 14, 2026 revision. Nothing here is quoted from memory or from secondary coverage. One typographic note: this site prints no dash characters, so where a source’s punctuation uses one, the quotation is broken around it or a colon stands in.