A peptide certificate of analysis should answer three questions first: what compound did the laboratory identify, what purity did its method measure, and how much material did the submitted sample contain? PeptidePrices calls these the core three-panel test: identity, purity and measured content. Endotoxin, microbial or sterility testing, heavy metals and fentanyl screening answer separate questions. Read the result beside its method, unit, sample description and report number.
Start with the certificate itself in the COA library, then use the COA checker to examine its issuer evidence. A large purity percentage is one result. The useful evidence is the complete, attributable report for the relevant research material and lot.
This guide is for research-use-only material and laboratory-document evaluation. It provides no dosing, preparation or medical advice. A research COA records analytical findings for submitted samples; it does not authorize human use.
What a COA actually records
A COA is a laboratory's written account of a particular analytical job. It connects a sample and client to methods, results and, where supplied, acceptance criteria. Good documentation lets another reader identify the tested material without guessing from a product photograph. The useful fields include the laboratory, report or accession number, client, product, nominal strength, lot, relevant dates, methods and results. Some reports also include a vial photograph, chromatogram, mass spectrum or verification code.
These fields serve different purposes. A vendor's product name describes the listing. The laboratory's sample name identifies what it received. A lot connects the report to a production group. An accession identifies the laboratory job. A search code opens the issuer's record. None should silently stand in for another. A report number printed in a batch field, for example, gives a reader no independent batch linkage.
Preserve the full document. A cropped purity screenshot may omit the second page with contamination results or the line naming a different client. If three vials were tested, retain the individual results and the stated mean. Three chromatograms can be replicates within one report, not three independent certificates. Similarly, a PDF and its image preview are two files representing the same evidence.
Read the core three-panel test first
| Panel | Main question | Read beside the result |
|---|---|---|
| Identity | Is the analytical finding consistent with the named compound? | Method, reference comparison and any spectrum |
| Purity | What share of the measured chromatographic signal belongs to the main component? | Detector, integration and separation conditions |
| Measured content | How much target analyte was measured? | Unit, calibration basis and sample or vial basis |
The core three-panel explainer walks through the distinctions with report examples. The three headings are questions, not necessarily three different instruments. A laboratory can use one chromatographic workflow for several outputs, while another combines chromatography and mass spectrometry.
Identity: method matters as much as “confirmed”
LC-MS combines liquid-chromatographic separation with mass spectrometry. A mass result consistent with the expected molecule supports its identification. A detailed assignment can also use retention behavior and fragment information. Matching an intact mass alone is narrower evidence than a full sequence investigation. Read what the report actually claims instead of expanding “LC-MS” into every possible structural test. Waters' peptide mass-confirmation application note shows how mass information and impurity profiling complement optical detection.
HPLC retention-time identity is a different comparison. It asks whether a sample peak appears where a reference is expected under specified conditions. That finding can be useful, but it should retain its method label. In the reviewed ZC Labs archive, ILS reports use retention-time identity while Freedom reports use LC-MS identity. A table that marks both simply “identity” describes coverage, not equivalent structural specificity.
The ILS Laboratories profile and Freedom Diagnostics profile provide context for those reporting patterns. On its own research-peptide methods page, ILS distinguishes its HPLC retention-time identity panel from custom sequence-confirmation work. Do not infer that every report from a laboratory used every instrument the laboratory owns.
Purity: a percentage of a defined analytical signal
HPLC-UV separates components and measures their ultraviolet response. A commonly reported chromatographic purity is the main peak's integrated area divided by the total integrated area considered by the method. Its denominator is the detected signal under those conditions, not the entire physical mass of the vial.
A clean-looking main peak still depends on separation. Two compounds that elute together may be difficult to distinguish in a UV trace. Components also differ in detector response. Water, salts and substances outside the method's detection or integration conditions are not automatically represented by the purity percentage. Waters' synthetic-peptide separation study illustrates why chromatographic resolution and mass detection belong together when assessing peptide impurities.
For a reader, the practical questions are concrete: is the chromatogram available, does the report name the detector, and is the stated purity linked to the correct sample? Do not turn 99% chromatographic purity into “99% of all vial contents are peptide.” The measured-content result addresses the amount of target analyte separately.
Measured content: read the amount independently
Measured content is often expressed in milligrams per submitted vial, but reports can also give concentration or another sample basis. Confirm the unit before comparing it with a label. A result in mg/mL cannot be compared directly with a nominal vial mass without the laboratory's volume basis. A bulk powder assay also cannot establish the fill amount of a finished vial.
Consider a hypothetical research vial labeled 10 mg, reported as 99% chromatographic purity and 8 mg measured content. The amount is 80% of the nominal label: 8 divided by 10, multiplied by 100. High purity and a low measured amount can occur together. This arithmetic is an illustration, not a measurement of a named vendor or a universal acceptance limit.
Review reported uncertainty, replicates and the stated specification before declaring a marginal difference a failure. A measured-content assay is also distinct from a biological activity assay. The FDA analytical-method validation guidance treats method performance as evidence that a procedure serves its intended analytical purpose. A method name without its intended measurement leaves an important question unanswered.
Extended panels answer separate questions
Endotoxin
Endotoxin testing examines bacterial endotoxin, rather than peptide identity or fill quantity. LAL refers to Limulus amebocyte lysate, a reagent used in established bacterial endotoxin methods. Read the result in endotoxin units and preserve the denominator: EU/mL, EU/mg and EU per vial are different expressions. A below-limit result should retain its inequality and reporting limit.
USP chapter 85 is the compendial reference for bacterial endotoxin testing. A “pass” needs a stated criterion and a method suitable for the sample. The endotoxin explainer explains report interpretation, including why a negative microbial result does not answer the endotoxin question.
Sterility or microbial screen
Look past the heading “sterility” to the method. USP chapter 71 concerns compendial sterility testing. A report labeled “Sterility (PCR)” describes a molecular microbial screen unless the issuer supplies evidence supporting a broader validated claim. PCR wording should remain visible when summarizing that result.
In ZC Labs' reviewed ILS reports, 23 of 24 include this PCR screen. That is a count of reported microbial screens, not 23 demonstrations of USP 71 testing. The sterility and microbial testing guide explains the distinction and how to read target scope, controls and sample linkage.
Heavy metals and elemental impurities
An ICP-MS panel measures specified elements. The report should identify those elements and give results, units and reporting limits. “Heavy metals: pass” is less informative than the element-by-element table. A four-element screen and a larger elemental panel should not be described as identical coverage.
ICH Q3D(R2) provides a risk-based framework for elemental impurities in drug products. It discusses sources such as manufacturing inputs, equipment and container systems. Its exposure framework is not a blanket approval of research-market material. The heavy-metal testing explainer focuses on interpreting ICP-MS findings without converting them into human-use guidance.
Fentanyl screen
A fentanyl result is a targeted screen. Retain the method, named target and detection threshold if the laboratory supplies them. “Not detected” means not detected within that assay's scope; it should not become a statement about all opioids, every possible contaminant or the absence of all foreign substances.
Freedom Diagnostics lists fentanyl screening separately from its core and other contamination tests. In PeptidePrices' retained Modern Research Peptides files, 86 of the 124 downloadable reports include a fentanyl result. That count describes documents with the panel, without suggesting how frequently contamination occurs across the market.
What the PeptidePrices library shows by laboratory
Our September 25, 2026 local audit counted 6,347 public certificate records after applying the site's existing rule for combining the fixed checker library with vendor archives. The scope includes current and historical records and all public verification states. It is a dated documentary snapshot, not the number of unique vials tested, unique laboratory jobs worldwide or current catalog matches.
The table below shows ten frequently represented laboratories. “Core” requires all three named core tests in the extractable fields. “Microbial” combines microbial and sterility labels once per record, including PCR screens. The counts are documented, extractable coverage, so an uncounted panel may be untranscribed. A zero is not evidence that a laboratory cannot perform a test.
| Laboratory | Records | Core | Endotoxin | Microbial | Metals | Fentanyl |
|---|---|---|---|---|---|---|
| Freedom Diagnostics | 2,120 | 1,592 | 637 | 276 | 111 | 202 |
| Janoshik | 855 | 590 | 159 | 144 | 86 | 0 |
| ILS Laboratories | 698 | 421 | 425 | 388 | 380 | 171 |
| Kovera Labs | 607 | 437 | 541 | 458 | 461 | 89 |
| Horizon Analytical | 353 | 168 | 195 | 0 | 0 | 0 |
| MDx BioAnalytical | 247 | 147 | 116 | 0 | 0 | 0 |
| Vanguard Laboratory | 218 | 50 | 34 | 36 | 38 | 0 |
| BioRegen | 202 | 113 | 50 | 0 | 0 | 0 |
| Chromate | 170 | 119 | 71 | 11 | 21 | 0 |
| BioViridian | 137 | 136 | 90 | 0 | 67 | 0 |
Within this captured evidence, extended panels occur frequently in Kovera and ILS records. Freedom has the largest report count in the table and includes both core-only and extended reports. The Janoshik records also span different requested panels. These patterns describe the certificates vendors published and PeptidePrices retained, not the complete service menu or output of any laboratory.
Use the Kovera Labs profile, Janoshik profile, Horizon Analytical profile and Chromate profile to investigate the issuer and verification route. Comparing laboratories by one aggregate percentage would mix different vendor requests, time periods and extraction completeness.
Vendor archives that make the distinctions concrete
Puratek Peptides: PeptidePrices verified 42 of 42 product-and-strength reports against official laboratory records, with 35 from ILS and seven from Freedom. All 42 report the core three-panel test; 41 include endotoxin, and 34 include heavy-metal and sterility-labeled panels. Analysis dates span February 1 through August 24, 2026. Those counts give readers a specific archive to inspect instead of an unsupported “fully tested” slogan.
ZC Labs: PeptidePrices matched 69 of 70 current and archived certificates to issuer records: 24 ILS and 45 Freedom matches. One archived Freedom report remained an issuer no-match. All 24 ILS reports include endotoxin; 23 include ICP-MS and a PCR microbial screen, and 20 include fentanyl screening. The archive dates span June 9 through September 23, 2026. The no-match remains outside the verified count.
Rebirth Labs: PeptidePrices verified all 119 reports admitted to the public archive against Freedom's records. The full recoverable collection contained 123 documents; four without retrievable issuer PDFs remain outside that public set. Among the 119 public records, extractable fields document 74 endotoxin and 46 microbial PCR panels. Rebirth is an example of why the public denominator and the retained audit denominator should both be named.
Modern Research Peptides: Of 125 archive records, 124 downloadable Freedom PDFs matched their accession, search code and printed MRP client. Those 124 contain core results; 82 include endotoxin, 36 microbial PCR, 27 ICP-MS elemental impurities and 86 fentanyl screening. One historical unavailable report is not verified. The metal and microbial totals differ, so neither panel can be inferred from the other.
Glacier Aminos: The reviewed public library has 270 records, including 245 exact issuer matches and 25 documentary-only records. The issuer matches comprise 113 Freedom, 112 Kovera, ten ILS, eight Chromate and two BioViridian records. Its public mix includes different laboratory methods and historical material. Use the current report attached to the exact listing when evaluating a particular strength.
Amino Club: PeptidePrices matched 102 issuer-verifiable records, comprising 93 ILS and nine Freedom records, in a 103-record archive. The evidence review describes ILS core, endotoxin, heavy-metal and sterility panels. Many stored rows summarize the report as “All 8 assays passed,” however, so the lab-wide automated table cannot allocate each unnamed assay. The 103 records represent 94 unique certificate documents. This is a concrete example of why record totals and document totals differ.
Refined BioLabs: All 25 public records carry issuer-verified status. The named panel fields include all three core tests plus endotoxin, heavy metals, a sterility label and fentanyl screening on all 25. Readers can inspect a consistently broad documented panel while still checking the microbial method on each report.
Validated Peptides: All 17 public reports are issuer matched and document the core three-panel test. Twelve carry extractable endotoxin, heavy-metal, sterility-labeled and fentanyl panels. This archive shows why a vendor-level testing description should preserve a 12-of-17 scope rather than imply that every report has the extended package.
Verify the report with the issuing laboratory
Start from the issuer's independently identified website or its linked PeptidePrices laboratory profile. Check the actual destination of a QR code: a vendor-hosted copy and a laboratory-hosted record are different sources. Enter the report number and any required access code exactly. Keep both the report and returned issuer record available for comparison.
Compare the printed client, report ID, product, strength, lot, dates and results field by field. A matching accession with a different client is a conflict needing explanation. A shared trade name or documented brand abbreviation can support attribution when the evidence establishes that relationship. Similar spelling alone cannot do that work.
Some portals return the entire document. Others confirm only that a number exists. Record which response you received. Existence-only confirmation cannot supply field comparisons that the portal never displayed. A portal outage, challenge page or unsupported laboratory lookup is an unavailable check. It is distinct from a completed lookup returning no match, and both differ from a record that directly contradicts the supplied file.
Red flags worth investigating
A wrong-client report can be authentic laboratory work without supporting the seller presenting it. Ask for documented attribution or a correctly attributed sample. A missing lot weakens the connection to a particular production batch. A cap color can be preserved as the laboratory's printed sample description, but it should not be silently promoted into a unique manufacturing lot.
A bulk raw-material report presented as a vial test crosses an evidence boundary. The raw material may have been tested before filling, packaging or later handling. Read the sample description and units to establish what stage the certificate covers. A vial photograph, submitted-unit count and per-vial results make that scope easier to evaluate.
A measured underfill deserves attention independently of purity. Check nominal amount, reported amount, sample basis and uncertainty; preserve the actual shortfall when established. For blends, look for component-specific amounts rather than treating one total as proof of every component. A cropped, incomplete or altered file should be checked against the full issuer copy before any result is repeated.
How PeptidePrices keeps those evidence boundaries
PeptidePrices retains complete reports and previews, records verifier inputs, and distinguishes exact matches from conflicts, no-matches, checked files and unavailable portals. The core checker and the expanded vendor archive remain separate evidence scopes. The public library brings them together for browsing without rewriting the original checker contract.
Listing controls use exact normalized vendor, canonical compound, strength and unit matches. Historical evidence remains searchable without automatically filling a current listing's missing certificate. For BPC-157, Tesamorelin, MOTS-c, GLP-2T and GLP-3R, open the report attached to the precise row being evaluated. A certificate for another strength is background evidence, not that row's test result.
The verification counts in this article describe PeptidePrices' existing documentary checks. We did not purchase, collect or retest the samples for this series. An issuer match establishes the documented report relationship; physical custody of a subsequently supplied vial is a separate evidence question.
Snapshot method and counting limits
The local audit uses the read-only library adapters, the archive database and generated archive mirror, the compact listing summary, and the vendor evidence records. The raw archive contains 6,708 rows; its generated discovery mirror describes 6,696 observations. Those include source observations and document parts. Neither is the merged public certificate count. The listing export represents 4,381 candidate records and is another separate scope. No counts from these overlapping layers are added together.
Panel totals use the existing named-test extraction from structured fields and retained OCR. The microbial column is a union, so a record labeled both microbial and sterility counts once. No compendial method is inferred from that column. Public certificate IDs are counted once, but different vendor records can represent the same underlying laboratory document. Vendor-specific reviewed ledgers supply the more precise document and verification counts above. Evidence and source checks are dated September 25, 2026.