Heavy metals testing is one of the few specification areas where a supplement brand owner cannot rely on a supplier's assurance alone. The limits are element-specific, the analytical methods have real detection floors, and the documentation trail is often thinner than the certificate of analysis suggests. This guide sets out what a contract manufacturer should be testing, against which thresholds, and what a buyer should request before placing a purchase order.
Which elements are in scope
"Heavy metals" is a colloquial term, not a regulatory one. In practice, four elements drive most specifications for dietary supplements: lead (Pb), cadmium (Cd), arsenic (As, reported as inorganic arsenic where the method allows), and mercury (Hg). These are the elements named in California's Proposition 65 listings and in most state-level and retailer-driven specifications.
Two further groups matter depending on the ingredient. First, the USP <232> elemental impurities list extends beyond the big four to include elements such as vanadium, nickel, cobalt, and thallium, classified by permitted daily exposure (PDE) and route. Second, raw materials from specific origins carry known risk profiles: marine sources (fish oils, algae, kelp) for arsenic and mercury; rice and cocoa for cadmium and inorganic arsenic; mineral clays and bentonite for lead; and botanical extracts for a mixed profile driven by soil and processing.
A specification that lists only "heavy metals, max 10 ppm" as a single summed figure is a legacy approach. It is not wrong, but it is blunt: it can pass a product with a high lead contribution offset by low cadmium, and it tells the buyer nothing about which elements were actually measured.
The limits that actually apply
There is no single federal limit for heavy metals in dietary supplements in the United States. Instead, a supplier and buyer typically work from a stack of overlapping requirements:
- 21 CFR 111 (current good manufacturing practice for dietary supplements) requires that finished product specifications include limits on contaminants and that you verify compliance. It does not set the numeric limit for lead or cadmium; it requires that you have one and that you test against it.
- 21 CFR 101.36 governs supplement facts labeling, including the declaration of lead where applicable, and interacts with how a product is positioned.
- USP <232> and <233> provide the elemental impurities framework and the analytical procedures (ICP-MS and ICP-OES). USP <232> sets PDE-based limits; for an oral product taken daily, the commonly cited PDEs are 5 µg/day for lead, 5 µg/day for cadmium, 15 µg/day for inorganic arsenic, and 3 µg/day for mercury (elemental). These are daily exposure limits, not concentration limits, so the permitted concentration depends on daily serving size.
- Proposition 65 in California sets safe harbor levels. For lead, the maximum allowable dose level (MADL) is 0.5 µg/day; for cadmium, 4.1 µg/day. These are far stricter than the USP PDEs and are the reason many brands adopt a Prop 65-compliant specification for the US market even when selling nationally.
- EU 1169/2011 and related EU contaminant regulations set the labeling and contaminant framework for product sold into Europe; the EU approach is generally contaminant-concentration based rather than PDE based, and the applicable levels differ by food category.
In practice, a defensible specification for a US-marketed supplement will state per-element concentration limits tied to the maximum daily serving, and will note whether the target is USP <232> PDE compliance, Prop 65 safe harbor, or both. Buyers should ask which framework the supplier's specification is built on, because "passes heavy metals" without a stated basis is not a specification.
Methods and detection limits
USP <233> describes two procedures: ICP-MS (Procedure 1) and ICP-OES (Procedure 2), with alternate procedures permitted under validation. Both are instrumental methods that report individual elements, which is what you want. The older colorimetric sulfide precipitation method reports a single summed result and has poor specificity; if a supplier's COA cites only "heavy metals (as Pb)" with a colorimetric method, that is a signal to ask for element-specific data.
Detection limits matter for two reasons. First, Prop 65 thresholds are low enough that a method with a limit of quantification around 1 ppm may not be able to demonstrate compliance for a small daily serving. Second, sample preparation can lose volatile arsenic or mercury if digestion is not controlled. A credible laboratory will report the method, the instrument, the limit of detection and limit of quantification, and the recovery or spike data for the batch.
Typical industry ranges, to be treated as indicative rather than universal: ICP-MS LOQs in the low parts-per-billion range for most of the big four; ICP-OES LOQs in the low parts-per-million range. Turnaround for a heavy metals panel at a commercial lab is commonly in the range of 5 to 10 business days for standard service, with expedited options shorter. These are typical ranges, not quoted prices or guaranteed timelines.
Documentation to request from a supplier
The certificate of analysis is the starting point, not the whole file. A buyer qualifying a supplier should request, at minimum:
- Element-specific COA for the finished lot, listing Pb, Cd, As, and Hg individually, with results, units, method reference, and specification limits.
- Raw material COAs for each ingredient in the formula, since finished-product testing alone can mask a high-contribution ingredient diluted by the batch.
- Method statement: which procedure (USP <233> Procedure 1 or 2, or an equivalent validated in-house method), instrument type, and LOQ.
- Third-party laboratory identity and accreditation, ideally ISO/IEC 17025 accredited for the specific test.
- Specification document showing the basis (USP <232> PDE, Prop 65, EU contaminant level) and the daily serving used to convert PDE to concentration.
- Stability data per ICH Q1A(R2) if the supplier claims heavy metals remain within specification over shelf life; this is relevant where packaging or ingredient interactions could concentrate an element.
- Supplier qualification records for the ingredient vendors, including any audit reports.
Where the product is imported, the importer's FSMA Foreign Supplier Verification Program (21 CFR 117) obligations and FDA Prior Notice (21 CFR 1005) filing should be visible in the file. A supplier that cannot produce these for an imported lot is a documentation risk regardless of the COA result.
Common failure modes
Most heavy metals problems a buyer encounters are not fraudulent; they are procedural. The recurring ones:
- Summed-result reporting. A single "heavy metals < 10 ppm" figure with no per-element breakdown, making it impossible to assess Prop 65 exposure.
- Method mismatch. A COA citing a colorimetric method while the specification references USP <233>. The result may be true but is not comparable.
- Wrong serving basis. A PDE converted to a concentration using a 1-gram serving when the label directs 3 grams per day. The limit silently triples.
- Raw material blind spot. Finished product passes because the risky ingredient is a small fraction of the formula, but a formulation change or a new supplier of that ingredient pushes the lot over.
- Stale or borrowed COAs. A COA from a different lot, or from a previous year, presented as current.
- No LOQ reported. A "not detected" result with no stated detection limit is not evidence of absence.
Questions to ask your supplier
Use these in a supplier questionnaire or during an audit. Vague answers are themselves informative.
- Which elements do you test for on every lot, and do you report them individually?
- Which method do you use, and what is the LOQ for each element?
- Is your specification based on USP <232> PDEs, Prop 65 safe harbor, or another framework, and what daily serving did you use to convert?
- Do you test raw materials as well as finished product, and can I see the raw material COAs for this formula?
- Which laboratory performs the testing, and is it accredited to ISO/IEC 17025 for this method?
- Can you provide stability data supporting the heavy metals specification through the declared shelf life?
- For imported lots, can you provide FSVP records and Prior Notice confirmation?
- What is your corrective action procedure if a lot exceeds specification?
If you are evaluating a new manufacturing partner, the depth of these answers is a reasonable proxy for overall quality systems; the same discipline shows up in certifications and audit readiness. Buyers sourcing dosage forms with a higher surface-area-to-mass ratio, such as capsule manufacturing, should confirm that the specification is set on the finished capsule weight and fill, not on the powder alone.
Frequently asked questions
Is there a federal limit for lead in dietary supplements in the US?
No single numeric federal limit exists for lead in dietary supplements. 21 CFR 111 requires that you establish contaminant specifications and verify compliance, but the numeric limit is set by the firm, typically by reference to USP <232> PDEs or California Prop 65 safe harbor levels. The buyer should confirm which basis the supplier used.
What is the difference between USP <232> and Prop 65 limits?
USP <232> sets permitted daily exposures for elemental impurities, including 5 µg/day for lead and 5 µg/day for cadmium for oral products. Prop 65 sets safe harbor MADLs that are stricter for lead (0.5 µg/day) and cadmium (4.1 µg/day). A product can meet USP <232> and still require a Prop 65 warning if sold in California.
Why does the COA say "heavy metals < 10 ppm" instead of listing each element?
That format reflects an older summed-result method, often colorimetric. It does not give you the per-element data needed to assess Prop 65 or to identify which raw material is contributing. Ask for an element-specific ICP-MS or ICP-OES result instead.
How long should heavy metals testing take?
Typical commercial laboratory turnaround for a four-element panel is in the range of 5 to 10 business days for standard service, with expedited options shorter. This is an industry range, not a guarantee; confirm with the specific laboratory and build the timeline into your lot release schedule.
Do I need heavy metals testing on every lot, or just at qualification?
Under 21 CFR 111 you must verify that finished product specifications are met, which in practice means lot-by-lot testing or a validated reduced testing program justified by supplier history and raw material risk. For high-risk ingredients such as marine oils, rice-derived materials, and mineral clays, lot-by-lot testing of the raw material is the safer default.
Heavy metals compliance is a documentation exercise as much as an analytical one. The buyer who asks for element-specific results, a stated specification basis, and the raw material trail behind the finished COA will catch most problems before they reach a label. Start with the supplier questionnaire above, request the full file for one lot, and compare what arrives against the specification the supplier claims to hold. That single exercise will tell you more than any brochure.




