Creatine Monohydrate COA: 7 Checks Beyond the Assay

Quick Answer

A creatine monohydrate certificate of analysis is useful only when its numbers describe the exact lot being purchased and the methods can answer the buyer’s questions. A headline such as “99% purity” does not, by itself, establish the assay basis, quantify synthesis-related impurities, show that the sample came from the shipment lot, or prove that the material meets the finished product’s requirements.

For a routine B2B qualification, compare seven things: identity and hydrate form; assay basis; method and sensitivity; creatinine, dicyandiamide and dihydrotriazine results; contaminant controls required for the destination; batch and laboratory traceability; and the written acceptance specification. Keep the listing, supplier’s typical specification, sample COA and shipment-lot COA separate. They are different evidence.

Why “99%” is not a complete answer

Creatine monohydrate contains one molecule of water of crystallization. The USP-NF preview defines creatine as the monohydrate and gives an assay of 98.0–102.0% for creatine calculated on the dried basis.[1] A supplier may instead describe a commercial grade as “99.9% creatine monohydrate,” use a different method, or quote a result corrected for water. Those values cannot be compared until the tested substance, calculation basis and method are known.

The distinction is practical. A dried-basis assay can exceed 100% without implying that the powder contains more than its total mass of creatine monohydrate: the result is corrected for the measured loss or water and reported on a defined basis. Conversely, a high principal-component assay does not prove that every relevant impurity is low. The COA must show what was measured and how. The buyer’s job is not to demand one universal number from every supplier. It is to define the intended grade and acceptance limits, then verify that the COA reports the same analytes, units and methods for each lot.

1. Confirm identity, form and grade

Start with the product name and chemical identity. “Creatine,” “creatine monohydrate” and anhydrous creatine are not interchangeable purchase descriptions. Check CAS number, molecular form, appearance and the supplier’s product code against the specification and purchase order. The monohydrate identity matters because water of crystallization is part of the defined compound, not automatically a defect or evidence of poor drying.[1]

The supplier should state whether the assay is expressed as creatine monohydrate or as creatine on a dried basis. If the listing says “food grade,” “USP grade” or “99.9%,” request the current grade specification and ask which monograph or internal standard the claim refers to. A marketplace label is not proof of compendial compliance. If the order requires a named compendial grade, state that requirement in the purchase specification and ask for the applicable test methods and lot results.

Particle size is a separate physical attribute. An 80-mesh and a 200-mesh grade may behave differently during conveying, blending, dust control or dispersion, but mesh alone does not establish purity, dissolution under a defined beverage process, or biological absorption. Keep the physical specification in its own line on the COA or TDS rather than treating it as a proxy for chemical quality.

2. Read the assay basis before comparing percentages

A COA should identify the assay as “as is,” “on dried basis,” “anhydrous basis,” or another precisely defined calculation. It should provide the result, acceptance range, units and calculation method. Where moisture or loss on drying is used for correction, the report should show the measured value or reference the linked test report.

Do not compare 99.0% as-is with 99.0% on dried basis as if they were equivalent. Nor should a buyer reject a result above 100% automatically if the compendial method reports the analyte on a dried basis. The USP definition is an example of why the basis matters: its preview specifies creatine content, calculated on the dried basis, separately from the monohydrate identity.[1]

Check the water-related test too. Loss on drying measures volatile matter removed under the stated conditions; it is not necessarily a water-specific measurement.[2] With a hydrate, a result near the expected crystal-water fraction may be consistent with the material’s identity, but a single LOD number cannot distinguish bound water from free moisture or other volatiles. If free moisture is a critical risk for flow, caking or shelf life, specify a suitable water method and limit instead of treating LOD and water content as synonyms.

A useful side-by-side comparison therefore records: analyte basis, assay method, water or LOD method, reported result, acceptance range and the product form to which each applies. If any one of those fields is missing, mark the result “not comparable” until the supplier clarifies it.

3. Ask what the analytical method can actually detect

The test name alone is not enough. “HPLC,” “USP method” or “tested in our lab” does not reveal whether the procedure separates creatine from the impurities of concern, uses appropriate standards, or is sensitive enough at the proposed limits.

The creatine monohydrate dossier submitted in FDA GRAS Notice 931 describes a validated in-house HPLC-UV method using external standards to determine creatine, creatinine, dicyandiamide and dihydrotriazine. It reports method-specific detection limits for those analytes.[3] This is useful evidence that a targeted chromatographic panel can measure more than the main assay. It is not proof that another supplier uses the same method, has the same detection limits, or meets that dossier’s specification.

For each result, ask for:

• The method number or a short method description, including detector and analyte.
• Whether the method is compendial, modified compendial or in-house, and whether it is verified or validated for this material.
• The reference standards used for creatinine, dicyandiamide and dihydrotriazine.
• The limit of detection (LOD) and limit of quantitation (LOQ) for each impurity.
• The result in a stated unit, not only “pass,” “conforms” or “not detected.”

“Not detected” means below a method’s detection threshold, not necessarily zero. “Pass” is meaningful only against a stated specification and method. For an impurity limit of 5 mg/kg, a method with an LOD of 10 mg/kg cannot demonstrate compliance at that limit. If the seller will not provide method details, request a third-party test on a representative sample before qualifying the supplier.

4. Review process-related organic impurities individually

For creatine monohydrate, the COA should address at least creatinine, dicyandiamide (DCD) and dihydro-1,3,5-triazine (DHT) when those analytes are relevant to the buyer’s intended use and specification. These are distinct compounds and should not be collapsed into a generic “total purity” statement.

A historical EFSA opinion discussed impurity limits for the specific creatine monohydrate evaluation before it, including creatinine, dicyandiamide and DHT. The values are tied to that assessment and its analytical context; they should not be copied as a universal limit for all products, jurisdictions or customers.[4] In a separate 2011 study, Moret and colleagues tested 33 creatine supplements using an optimized HPLC method. Creatinine was above 100 mg/kg in 44% of the samples; the authors also reported detectable DHT and some dicyandiamide results above 50 mg/kg.[5] The sample set is small and old, so it cannot describe today’s market. It does show why the main assay is not a substitute for targeted impurity results.

Ask the supplier to list each analyte, its specification, actual lot result, unit and method. If a report gives only “total impurities,” ask which compounds are included, how the total is calculated and whether the method detects each one at the agreed limit. A single HPLC assay may not quantify all impurities unless the method is designed and validated to do so.

Keep a distinction between a supplier’s typical specification and the current batch’s result. A technical sheet with “creatinine ≤50 mg/kg” is a proposed limit. It is not evidence that the lot offered to you measured 50 mg/kg or lower. The lot COA should name the batch and show the actual result, or clearly identify a permitted skip-testing or reduced-testing rationale in the quality agreement.

5. Match metals and microbiology to the product and market

Heavy-metal and microbiological tests belong in the quality review, but their exact panel and limits should come from the finished product’s risk assessment, customer requirements and destination-market rules. Do not accept a combined “heavy metals: 10 ppm” line if the buyer needs individual lead, arsenic, cadmium and mercury results. Ask for element-specific limits, actual values, units, method and lab report where required.

The dietary-supplement CGMP rule in 21 CFR 111.70 requires manufacturers to establish component identity specifications, specifications needed to ensure the finished supplement’s purity, strength and composition, and limits for contamination that could adulterate the finished batch.[6] The rule does not set one universal creatine monohydrate COA template or a single impurity limit for every use. The responsible manufacturer must translate the product and market requirements into documented specifications.

6. Verify that the certificate belongs to the offered lot

A clean-looking PDF can still be the wrong document. Compare the COA with the quote, product code, specification, sample label and shipping paperwork. At minimum, the certificate should identify the manufacturer, material, batch or lot number, test date, report date, results, methods, specification limits and an authorized reviewer or issuing laboratory.

Check the chronology. The manufacturing and testing dates should make sense for the batch, and the certificate should not be a generic sample with no lot identifier. If a broker provides the document, ask who manufactured the material and who tested it. A supplier’s name on the invoice does not prove that it made the powder or generated the laboratory data.

For an external report, confirm the laboratory name and accreditation status. ISO/IEC 17025 accreditation is useful only when the laboratory’s scope covers the relevant method or analyte; the logo alone does not show that. If the laboratory’s report number can be verified directly, do so through the lab’s published channel. Preserve the source PDF and the verification record in the incoming-lot file.

For first qualification, compare a pre-shipment sample with the final lot documentation, then define whether the sample is representative and how retained samples will be handled. If the supplier changes the manufacturing site, raw-material route, specification or test method, require advance notice and assess whether requalification is needed.

For microbiology, verify which organisms or counts are tested, the sample size, method and acceptance limit. A supplier’s general certificate or facility certificate does not replace a lot result when the buyer’s specification requires testing. Conversely, do not add an arbitrary panel without a risk or regulatory reason; agree the required tests before purchasing so the supplier and receiving lab work to the same release criteria.

The product listing and the supplier storefront are linked in the sourcing card below. The supplier’s public product information should be reconciled against the exact product code, purchase specification and lot documents before any order.

7. Turn the COA into a release decision

Before the first order, write the acceptance specification into the purchase order or quality agreement. Include the identity and form, assay basis and range, water or LOD requirement, individual organic impurities, relevant contaminant tests, methods, units, sampling plan and document requirements. Identify which values are release-critical and what happens if a result is missing, out of specification or disputed. A practical review can use three outcomes:

• Accept for the defined use: identity matches; required tests are present; results meet written limits; the lot and methods are traceable.
• Hold for clarification or testing: the analyte or method is unclear, the COA is not lot-specific, or the method is too insensitive to evaluate the buyer’s limit.
• Reject or investigate: identity conflicts, a required result exceeds the agreed limit, the lot cannot be traced, or a material document discrepancy remains unresolved.

This is a buyer’s release framework, not a universal compendial decision tree. The qualified person responsible for the finished product should set the actual limits and disposition procedure. Where results conflict, quarantine the lot, review sampling and method suitability, and use an independent laboratory before changing the specification or accepting the material.

For repeated purchasing, convert a successful first qualification into a controlled routine rather than relying on the first COA forever. Set the incoming test frequency by supplier history, process risk, lot size and the consequences of a failure. A buyer might test every early lot for the full panel, then reduce selected tests only after documented evidence supports that decision and applicable rules permit it. Keep identity verification and traceability intact even when some contaminant tests are reduced. Reopen qualification after a source, site, process, test method or specification change, and when a trend approaches an acceptance limit. A passing result on one lot does not establish that later lots will match; the purchase agreement should define the change-notification and escalation steps before a problem occurs.

The main purchasing lesson is simple: compare evidence in the same units and on the same basis. A 99% label, a general specification sheet and a shipment COA answer different questions. Ask for all three, connect them to the same product code and lot, and release only against limits your own product and market require.

Sourcing lead for evaluation

An Alibaba listing from Foodchem International Corporation identifies creatine monohydrate, CAS 6020-87-7, lists 80/200 mesh and a stated minimum assay, and displayed 18,100 sold when checked on October 21, 2026. The supplier profile displayed 20 years on Alibaba. These are marketplace signals, not independent proof of repeat purchase, manufacturing quality or the current lot’s compliance. The listing does not display a lot-specific result for creatinine, DCD or DHT. Treat it as a quote and sample lead; request the current specification and batch COA before qualification. The product page also presents different particle sizes as improving absorption; do not use that claim as a raw-material quality conclusion without independent evidence.

References

1. United States Pharmacopeia. Creatine, USP-NF Preview. Identity and assay definition for creatine monohydrate, including dried-basis assay. https://doi.usp.org/USPNF/USPNF_M20457_02_01.html.
2. United States Pharmacopeia. General Chapter <731>, Loss on Drying. https://doi.usp.org/USPNF/USPNF_M99510_02_01.html.
3. U.S. Food and Drug Administration. GRAS Notice No. 931: Creatine Monohydrate. Notifier-submitted dossier describing identity, specifications and a validated HPLC-UV method for a specific proposed ingredient. https://www.fda.gov/media/143525/download.
4. EFSA Panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food. Opinion related to creatine monohydrate for use in foods for particular nutritional uses. EFSA Journal. 2004;2(4):36. https://doi.org/10.2903/j.efsa.2004.36.
5. Moret S, Prevarin A, Tubaro F. Levels of creatine, organic contaminants and heavy metals in creatine dietary supplements. Food Chemistry. 2011;126(3):1232–1238. https://doi.org/10.1016/j.foodchem.2010.12.028.
6. U.S. Electronic Code of Federal Regulations. 21 CFR 111.70, What specifications must you establish? https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-111/subpart-E/section-111.70.