Dry-Weight Basis (Potency)

Every potency percentage is a fraction: mass (or activity) of the analyte of interest divided by the total mass of the sample as it was weighed. The three common reporting bases differ only in what goes into that denominator, and confusing them produces numbers that look like a discrepancy but are actually a definitional mismatch.

Dry-weight basis and anhydrous basis are frequently used loosely as synonyms, but they are conceptually distinct. Anhydrous basis corrects for a specific, chemically defined water of hydration that is part of a crystal structure and can be characterized precisely (e.g. by loss-on-drying at a temperature that drives off hydrate water without decomposing the compound). Dry-weight basis corrects for bulk, variable moisture content in a heterogeneous natural material like plant matter, where the 'water' being removed has no fixed stoichiometric relationship to the analyte.

The moisture fraction used in a dry-weight conversion is itself a measured value, and the method used to measure it matters — the two dominant compendial methods measure different things and are not interchangeable for materials where volatiles other than water contribute to weight loss.

USP <731> is the default for most solid dosage-form components and is simple, cheap, and widely validated, but it is a poor choice for materials where a meaningful fraction of the weight loss on heating is not water — essential-oil-bearing botanicals, materials with residual ethanol from extraction, or heat-labile actives that decompose before moisture fully drives off. USP <921> Karl Fischer is specific to water and is the method of choice whenever LOD's non-specificity would bias the dry-weight conversion, or whenever the material's specification explicitly calls for water content rather than generic loss on drying.

The core relationship converting between bases uses the moisture fraction (m, expressed as a decimal) determined by LOD or Karl Fischer on a representative, matched sample:

1. Determine moisture fraction m (e.g. 0.095 for 9.5% moisture) on a sample from the same lot and, ideally, the same sub-sample as the potency assay.
2. As-is result → Dry-weight result: DWB = As-is ÷ (1 − m).
3. Dry-weight result → As-is result: As-is = DWB × (1 − m).
4. Anhydrous conversion for a hydrate salt uses the same formula with m as the known or measured water-of-hydration fraction, often derived from the molecular weight ratio of hydrate to anhydrous form rather than from an LOD measurement.

The last row is included deliberately: it illustrates the most common conversion mistake, which is applying a dry-weight/anhydrous conversion to a result that was already reported on that basis, effectively double-correcting for moisture and producing a nonsensical result over 100%. Any conversion output above 100% (for a percentage-of-total-mass measurement) is a strong signal that the basis of the starting result was misidentified.

Dry-weight basis is not merely a convention in the hemp/cannabis space — it is a statutory testing requirement. Under 7 CFR Part 990 (the USDA Domestic Hemp Production Program), the legal definition of hemp (≤0.3% total THC) is explicitly measured on a dry-weight basis, meaning the moisture content of the flower sample must be determined and the total THC result mathematically corrected to the dry-weight equivalent before the 0.3% threshold is applied. A sample that reads 0.27% total THC as-is at 15% moisture converts to 0.27 ÷ 0.85 = 0.318% dry-weight — crossing the federal threshold purely because of the basis conversion, with no change in the actual as-received material.

State cannabis programs vary in whether potency is reported as-is or dry-weight for finished, dispensable product, but nearly all require dry-weight basis for the compliance boundary between 'hemp' and 'marijuana' under the federal definition, because that boundary is defined in statute on a dry-weight basis. Laboratories testing for USDA hemp compliance must therefore run a moisture determination alongside every cannabinoid potency test and apply the conversion before reporting a compliance result.

USP <2030>, Supplemental Information for Articles of Botanical Origin, and individual USP-NF botanical monographs typically specify both a loss-on-drying limit (an upper bound on acceptable moisture, since excess moisture drives microbial growth and degradation) and a marker-compound assay, and the monograph states explicitly whether the assay result is reported as-is or on a dried basis. Under 21 CFR 111.75, dietary supplement manufacturers must establish component specifications for identity, purity, strength, and composition — and the basis convention for the strength specification must be documented as part of that specification, not left implicit.

Because botanical raw materials are agricultural commodities with genuinely variable moisture from harvest to harvest and lot to lot, the dry-weight basis conversion is what allows a fair, apples-to-apples comparison of marker-compound content across lots that may have been dried, stored, or shipped under different humidity conditions. Reporting strictly as-is would make lot-to-lot potency comparisons meaningless, since a wetter lot would always appear artificially low in marker content relative to a drier lot of genuinely identical composition.

The single most common basis-related failure is not a calculation error — it is a missing or ambiguous basis declaration on the certificate of analysis itself. A CoA that reports '9.2% marker compound' without stating as-is or dry-weight forces the receiving quality unit to guess, and different operators or shifts may guess differently, producing inconsistent dispense targets for genuinely identical material.

1. Supplier CoA omits the basis declaration entirely — the receiving lab must contact the supplier or, if unresolved, treat the lot as non-conforming pending clarification rather than assume a basis.
2. Supplier changes basis convention between lots without flagging it — a supplier switches labs or testing methods and the new lab defaults to dry-weight while the prior lab reported as-is; potency appears to jump lot-to-lot with no compositional change.
3. In-house re-assay reports on a different basis than the incoming CoA — the receiving lab's standard method uses LOD-corrected dry-weight while the supplier reports as-is, and the discrepancy is investigated as an OOS before anyone checks the basis.
4. Specification silent on basis — the written specification states a potency range with no basis annotation, so conformance decisions are made inconsistently across reviewers.

Basis mismatch does not only distort the reported potency percentage — it propagates directly into the dispensed mass and, from there, into yield reconciliation. If the formula's target assumes a dry-weight potency but the dispense engine applies the assay-adjusted charge using the as-is number for the same lot, the operator will weigh out more material than the formula requires (because the as-is number reads lower than the dry-weight number for any moist material), and the finished batch will run over target on that component while another component may be short if total batch mass is constrained.

Under 21 CFR 211.103, yield calculations must be reconciled against the adjusted theoretical yield, and an adjusted theoretical built on a basis-mismatched potency number is not a valid adjusted theoretical — it will generate a yield variance that traces, on investigation, not to a process failure but to a units-of-measure error, wasting investigation time and potentially masking a real process deviation underneath the noise.