Is my cleaning acceptance limit defensible, or do I need health-based limits instead of 10 ppm?

This is United States federal law for dietary supplements: 21 CFR part 111. You are subject to it if you manufacture, package, label or hold a dietary supplement, including one you manufacture that somebody else packages or labels, and one imported or offered for import (21 CFR 111.1(a), 111.1(a)(1), 111.1(a)(2)). The one exception in that section is narrow: the requirements that pertain to holding do not apply where you hold at a retail establishment for the sole purpose of direct retail sale to individual consumers, and a retail establishment does not include a warehouse or other storage facility for a retailer, or a warehouse or other storage facility that sells directly to individual consumers (111.1(b)). Part 111 is also a floor rather than the whole of it: you must comply with other applicable statutory provisions and regulations under the Act related to dietary supplements as well (111.5).

Read this before you act on anything written about 10 ppm or health-based limits. Almost all of it was written for a reader you may not be, and the applicability clauses are worth reading first.

• The health-based exposure limit is European guidance for medicines. It comes from the European Medicines Agency’s guideline on setting health based exposure limits for use in risk identification in the manufacture of different medicinal products in shared facilities (EMA/CHMP/CVMP/SWP/169430/2012, adopted 20 November 2014, in effect from 1 June 2015). Its stated scope is the safety of human patients and target animals exposed to residual active substances via medicinal products, and of consumers exposed through food of animal origin from treated food-producing animals. Its legal basis section lists four instruments to be read in conjunction with it, the first being EudraLex Volume 4, the European GMP guidelines, chapters 3 and 5. Standing: agency guidance of a European regulator, for medicinal products. It is not United States law and it does not reach a dietary supplement.
• The United States drug rule does not reach you either. 21 CFR part 211 contains the minimum current good manufacturing practice for preparation of drug products, excluding positron emission tomography drugs and medical gases, for administration to humans or animals (21 CFR 211.1(a)). And it carries a non-enforcement provision that reaches a further group of readers: pending consideration of a proposed exemption published in 1978, the requirements of part 211 are not enforced for OTC drug products if the products and all their ingredients are ordinarily marketed and consumed as human foods and those products may also fall within the legal definition of drugs by virtue of their intended use; and parts 110 and 117, and where applicable parts 113 through 129, are applied to those products instead (211.1(c)).
• If you reached for the FSMA preventive controls rule instead, check two exemptions. Subparts C and G of part 117 do not apply to any facility with regard to the manufacturing, processing, packaging or holding of a dietary supplement that is in compliance with part 111 and section 761 of the Act (21 CFR 117.5(e)). And except as provided by subpart E, those subparts do not apply to a qualified facility, which carries the modified requirements of 117.201 instead (117.5(a)). Neither exemption touches subpart B, and neither is a reason to skip part 111.

So if you make a dietary supplement, the whole 10 ppm-versus-HBEL argument is happening in rules you are outside of. That does not make the question academic. It changes the answer.

On this page: Where the two numbers come from · What part 111 actually asks of your limit · What the inspection record shows · What makes a limit defensible · What this does not answer

Where the two numbers come from

Neither figure appears in the regulation you are measured against. Across the whole of 21 CFR part 111, the words residue, carryover, ppm and health-based do not appear at all, and neither does validation. The rule regulates cleaning without ever naming a number for it.

The 10 ppm default is an industry convention. It is carried between firms, consultants and template packs, and it is not a federal requirement for a dietary supplement. It is also not stated in the document usually said to have replaced it: the EMA guideline does not use the figure anywhere in its text. What that guideline says about existing practice is narrower and more useful — that a variety of approaches are taken to establish carry-over limits for cleaning validation and often do not take account of the available pharmacological and toxicological data.

The health-based limit is a derivation, not a number. The guideline recommends the permitted daily exposure method described in Appendix 3 of ICH Q3C(R4) and Appendix 3 of VICH GL 18: a no-observed-adverse-effect level for the critical effect, adjusted for body weight, divided by five factors that account for extrapolation between species, variability between individuals, short study duration, severe toxicity, and the absence of a no-effect level. Where several critical effects produce more than one value, the guideline says the lowest is usually used, with a justification for the choice.

Three qualifiers travel with that method and are routinely dropped when it is repeated. The guideline states in its own executive summary that deviation from the approach it sets out could be accepted if adequately justified, and again in the calculation section that other approaches to determining health-based exposure limits could be considered acceptable if adequately and scientifically justified. It sets a different route for active substances with a genotoxic potential and no discernible threshold, where a limit dose of 1.5 micrograms per person per day may be applied. And for therapeutic macromolecules and peptides it says that because cleaning typically exposes surfaces to pH extremes or heat that degrade and inactivate protein-based products, determining health-based exposure limits using the limits of the active and intact product may not be required, adding that where other potential routes of cross-contamination exist the risks posed should be considered case by case.

Read as it is written, the guideline is an argument that a limit should be derived from what is known about the substance rather than from convention. That argument survives the trip across the Atlantic. The obligation does not.

What part 111 actually asks of your limit

Start with the cleaning obligation itself, because it is more conditional than it is usually quoted as being. You must maintain, clean and sanitize, as necessary, all equipment, utensils and any other contact surfaces used to manufacture, package, label or hold components or dietary supplements (21 CFR 111.27(d)). Equipment must be taken apart as necessary for thorough maintenance, cleaning and sanitizing (111.27(d)(1)). Contact surfaces used for low-moisture components or supplements must be in a dry and sanitary condition when in use, and where they are wet-cleaned they must be sanitized when necessary and thoroughly dried before subsequent use (111.27(d)(2)). A separate paragraph applies only if you use wet processing during manufacturing, and it is the one that reaches interruptions and consecutive batches: clean and sanitize all contact surfaces, as necessary, to protect against the introduction of microorganisms into components or dietary supplements, and, when cleaning and sanitizing is necessary, clean and sanitize all contact surfaces before use and after any interruption during which the contact surface may have become contaminated. Where surfaces are used in a continuous production operation or in consecutive operations involving different batches of the same dietary supplement, adequately clean and sanitize as necessary (111.27(d)(3)). Cleaning compounds and sanitizing agents must be adequate for their intended use and safe under their conditions of use (111.27(d)(6)).

Notice what that vocabulary is about. The rule’s own definition of sanitize is a microbiological one: to adequately treat cleaned equipment, containers, utensils or any other cleaned contact surface by a process effective in destroying vegetative cells of microorganisms of public health significance and substantially reducing numbers of other microorganisms, without adversely affecting the product or its safety for the consumer (111.3). A contact surface is any surface that contacts a component or dietary supplement, and those surfaces from which drainage onto the component or supplement, or onto surfaces that contact it, occurs during the normal course of operations (111.3). Part 111 does not carry a chemical carryover vocabulary. That is why looking in it for a residue number returns nothing.

The written-procedure and record obligations sit beside it. Under subpart D you must establish and follow written procedures for fulfilling the requirements of that subpart, including for maintaining, cleaning and sanitizing, as necessary, all equipment, utensils and any other contact surfaces (111.25, 111.25(c)). Those written procedures are themselves a record you must make and keep (111.35(b)(1)(iii)), alongside documentation, in individual equipment logs, of the date of the use, maintenance, cleaning and sanitizing of equipment — unless that documentation is kept with the batch record (111.35(b)(2)).

Now the part your number actually lives in. You must establish a specification for any point, step or stage in the manufacturing process where control is necessary to ensure the quality of the dietary supplement and that it is packaged and labeled as specified in the master manufacturing record (111.70(a)). For in-process production you must establish in-process specifications for any point, step or stage in the master manufacturing record where control is necessary to help ensure that specifications are met for the identity, purity, strength and composition of the dietary supplements and, as necessary, for limits on those types of contamination that may adulterate or may lead to adulteration of the finished batch (111.70(c)(1)). A residue acceptance limit on shared equipment is one of those limits when control at that point is necessary.

Then the two paragraphs that answer the question in the heading. You must provide adequate documentation of your basis for why meeting the in-process specifications, in combination with meeting component specifications, will help ensure that the finished specifications and the contamination limits are met (111.70(c)(2)). And quality control personnel must review and approve that documentation (111.70(c)(3)). One short sentence carries the rest: you must determine whether the specifications you establish under 111.70 are met (111.73).

Around all of it sits the general obligation to take all the necessary precautions during manufacture to prevent contamination of components or supplements, an obligation whose listed precautions include performing mechanical manufacturing steps such as grinding, blending and sifting by any effective means to protect against contamination — cleaning and sanitizing contact surfaces being given as an example of such a means rather than as the prescribed one (111.365, 111.365(h), 111.365(h)(1)), and the requirement to conduct all manufacturing operations in accordance with adequate sanitation principles (111.360). Quality itself is defined as the dietary supplement consistently meeting the established specifications for identity, purity, strength and composition, and limits on contaminants, and having been manufactured, packaged, labeled and held under conditions to prevent adulteration under section 402(a)(1), (a)(2), (a)(3) and (a)(4) of the Act (111.3).

So the rule never gives you a number and never asks you to borrow one. It asks three things of whatever number you write: that you established it because control at that point was necessary, that you wrote down why meeting it does the job, and that your quality control personnel reviewed and approved that writing. A limit derived from a health-based exposure assessment can meet all three. So can a limit derived another way. What a default carried in from somewhere else has trouble with is the second, because the derivation is the part that did not travel with the figure.

That is a general pattern, not a cleaning one. Numbers in routine use often turn out to have no derivation behind them, traceable to convention, to recollection, or to a figure its own setters called arbitrary. And where a limit is genuinely traced, it is commonly derived backwards from what a process has already shown it can achieve rather than from what the requirement demands. Cleaning is where it lands hardest, because the level at which a contaminant counts is set by what the next product can tolerate, not by any general housekeeping figure — which is why a facility has to derive its own values from the equipment it runs and the products it runs on it, rather than adopt one. The same reasoning applied to product limits is the specifications question next door. A count of validation runs carries the same problem, and that one is how many PPQ batches you can actually defend.

What the inspection record shows

FDA publishes the observations its investigators write during inspections. Across 23,440 observations citing 21 CFR part 111, in fiscal years 2009 to 2026, not one uses the words residue, carryover, ppm or health-based. The federal inspection record on supplement cleaning is not an argument about the number.

What it does contain, on the clauses above:

The 132 observations under 111.25(c) separate, because FDA’s own wording separates them: 60 read that the written procedures were not established, 40 that they were not followed, 31 that they were neither established nor followed, and one row is malformed in the source. Read from the inspection side, the same clauses are where a cleaning validation is likely to have a hole.

What this counts, and what it does not. These are counts of clause citations in inspection observations, one row per citation, read from FDA’s published inspection observation records on August 18, 2026. Fiscal year 2026 was still open at that date. They are not counts of firms, inspections, warning letters or recalls. A citation recording that a procedure was not established is evidence that a document was absent; it is not evidence about the soundness of the procedures that do exist, and only the 111.25(c) row separates the two, because the 111.27(d) wording does not distinguish them. The 111.70(c)(1) and (c)(2) figures cover every in-process specification, so neither is a count of residue-limit findings. The clause selection is ours and is not exhaustive: a cleaning problem can be written up under clauses not listed here.

What makes a limit defensible

Four things are true of a limit that holds up when somebody reads it, and you can check all four without buying anything.

1. You can name what it was derived from. A health-based exposure assessment for the worst-case active. A toxicological reference for the substance. Your own contamination limit for the next product, worked back to the equipment. Any of those is a source. A figure that arrived with a template is not one.
2. The derivation is written down, not remembered. That written basis is what 111.70(c)(2) asks for, and it is what a borrowed number arrives without.
3. Quality control personnel reviewed and approved it (111.70(c)(3)). An approval on the protocol is not an approval of the basis document unless the basis document is in it.
4. A method can actually reach it. You must determine whether the specification is met (111.73), and a limit set below what your swab or rinse method can reliably measure is one you cannot determine either way. The method’s capability is a floor under every number on the page.

If you are certified rather than only inspected, the scheme you certify to is a second reader with its own expectations. NSF/ANSI 455-2, the dietary supplement GMP scheme, addresses cleaning verification and validation records at clause 4.4.22.4, and treats them as a recommendation conditioned on where they are required rather than as a fixed numeric limit. Your certification body’s own reading of your scope governs there, not ours.

What this does not answer

One question, one article. This one covers where a cleaning acceptance limit legitimately comes from and what part 111 asks of it. It does not cover how many runs prove a cleaning process, how to pick the worst-case product and group the rest, how to design or apply a swab recovery study, whether a shared line should be dedicated instead, or what any authority outside the United States requires. Each of those is a separate question with a separate answer, and compressing one into a paragraph here is what would make the paragraph wrong.