ICP-MS vs. ICP-OES: How to Choose the Right Heavy Metal Method for Your Product

There’s a question we get at least once a week from supplement and cosmetics brands: “Do you test with ICP-MS?” The assumption underneath it is that ICP-MS is the gold standard and ICP-OES is the lesser option — the method a less serious lab uses. That framing costs brands real money and, in some cases, generates data that’s harder to defend to a regulator, not easier.

The right answer isn’t ICP-MS or ICP-OES. It’s the method that matches your matrix, your regulatory threshold, and your product claim. Getting that choice wrong in either direction — over-specifying or under-specifying — shows up in your COA whether you realize it or not.

The Same Goal, Very Different Physics

Both instruments measure elemental concentrations — lead, arsenic, cadmium, mercury, chromium, and dozens of others — by introducing a prepared sample into an argon plasma burning at roughly 6,000–10,000°K. That’s where the similarity ends.

ICP-OES (Inductively Coupled Plasma Optical Emission Spectrometry) works by reading the light emitted when excited atoms return to their ground state. Every element emits at characteristic wavelengths, and the instrument measures emission intensity to calculate concentration. It’s fast, mechanically robust, and cost-effective for elements present at the parts-per-billion to parts-per-million level. A well-run ICP-OES can analyze 30 or more elements simultaneously in under two minutes per sample.

ICP-MS (Inductively Coupled Plasma Mass Spectrometry) takes a different route. Ions are extracted from the plasma, separated by mass-to-charge ratio in a quadrupole or sector-field analyzer, and individually counted. That detection mechanism pushes sensitivity into the sub-parts-per-trillion range for many elements. It also introduces complexity — polyatomic spectral interferences, matrix suppression effects, and instrument drift that require careful, ongoing method validation to manage.

Neither approach is categorically better. They solve different analytical problems, and a reputable cosmetic testing laboratory or supplement testing lab should be running both.

Detection Limits: Where the Numbers Actually Matter

The performance gap between the two methods is quantifiable, and it matters most when your regulatory compliance window is narrow.

For a standard ICP-OES method in aqueous solution, typical instrument detection limits (IDLs) run approximately 0.1 μg/L for lead, 0.5 μg/L for arsenic, and 0.05 μg/L for cadmium. After digestion and dilution in a real product matrix, your working method detection limit (MDL) in the actual sample is often 10–50× higher than those IDLs.

For ICP-MS, those IDLs drop to the 0.001–0.005 μg/L range for the same elements. A well-validated ICP-MS method routinely achieves MDLs below 0.01 μg/L in finished product matrices — a 10- to 100-fold improvement depending on the element.

That gap becomes regulatory reality fast. California’s Proposition 65 sets the reproductive toxicity MADL for lead at 0.5 μg/day. If your supplement serves a 2-gram dose, you need a defensible detection limit of roughly 0.25 μg/g or better to make a compliance statement with confidence. ICP-OES can struggle to deliver that consistently. ICP-MS makes the margin comfortable.

USP <232>, governing elemental impurities in dietary supplements and pharmaceutical products, sets PDEs (Permitted Daily Exposures) at 5 μg/day for lead, 2 μg/day for cadmium, and 15 μg/day for arsenic for oral route products. Depending on your labeled serving size and daily dose, reaching those thresholds in the sample requires sub-ppb detection — territory where ICP-MS is the technically sound choice.

For cosmetics, FDA’s guidance referencing the ICCR framework cites a lead limit of 10 μg/g for cosmetic products. At that concentration, either method works reliably. But if you’re also evaluating for mercury (FDA has pursued enforcement actions against skin-lightening products at 1 μg/g) or if you’re selling into a market that demands sub-ppm arsenic data, ICP-MS earns its cost.

When ICP-OES Actually Outperforms Its More Expensive Sibling

Here’s the part that doesn’t get said enough: in high-matrix samples, ICP-OES frequently produces cleaner data than ICP-MS, and routing everything through ICP-MS without thinking about matrix effects can actively degrade your results.

Consider a calcium carbonate supplement at a 1,000 mg daily dose. Each analytical portion carries a massive calcium load into the instrument. In an ICP-MS, that calcium generates polyatomic interferences: ArCa⁺ at mass 56 suppresses accurate iron quantitation; CaO⁺ at mass 57 interferes with manganese; Ca₂⁺ contributes to barium signals. Modern instruments use collision-reaction cells (CRC) and kinetic energy discrimination to manage these, but every correction step adds measurement uncertainty. Labs that don’t actively validate for these interferences in calcium-heavy matrices hand clients COAs with quietly inflated chromium, manganese, or barium values — and no one catches it until a regulatory review.

ICP-OES handles high-calcium matrices with fewer artifacts because the optical detection system is inherently less susceptible to the ion-based interference mechanisms that trouble mass spectrometry. For high-mineral botanical extracts, marine-derived ingredients, calcium-based antacids, and electrolyte powders, we often run ICP-OES as the primary method and add ICP-MS selectively for the elements where sub-ppb limits are required — arsenic, mercury, and lead for Prop 65, for instance.

For nutritional label claims on minerals — iron, zinc, magnesium, selenium content, copper — ICP-OES provides perfectly adequate quantitation at the concentrations where those elements appear in finished products. Paying for ICP-MS sensitivity to confirm that your magnesium supplement contains 300 mg of magnesium is, bluntly, wasted budget.

A Practical Decision Framework by Product Category

Rather than prescribing a single method universally, here’s how method selection actually maps to product type in practice:

Dietary supplements (oral products) Reference USP <233> for method suitability requirements and USP <232> for acceptance criteria. For California Prop 65 compliance — particularly if you’re shipping into the state or selling on Amazon to California residents — ICP-MS is typically required for lead and inorganic arsenic due to the low daily intake thresholds. For nutritional mineral panels and mid-level element screening, ICP-OES handles most analytes reliably.

Cosmetics and personal care products The MoCRA framework (signed into law as part of the FD&C Act amendments) doesn’t mandate a specific analytical instrument, but it does require that safety data be scientifically sound and maintainable in records available to FDA. For a full heavy metal screen in a lip product, pressed powder, or skin cream — lead, arsenic, cadmium, mercury, chromium, nickel, antimony — ICP-MS is the defensible choice. The incidental ingestion exposure pathway for lip products in particular creates a daily intake concern that ICP-OES margins don’t comfortably cover for lead.

Food products FDA’s action levels and guidance limits for heavy metals in most food commodities sit at the ppm level, which ICP-OES covers reliably. Grains, nuts, proteins, spices — standard food safety screening for these matrices is well within ICP-OES performance. The exception is infant formula and baby food. FDA’s draft guidance on toxic elements in baby food and infant formula targets lead at 10 μg/kg (ppb) in processed fruit, vegetables, and grain-based products. Getting defensible quantitation at that level requires ICP-MS.

Raw material qualification If your finished product specification demands ICP-MS-level detection, qualifying the incoming raw material with ICP-OES creates a logical gap in your supply chain documentation. Work backward from your finished product limits when specifying incoming material specs, and make sure the method tier matches.

What Your Lab’s ISO 17025 Scope Is Actually Telling You

When evaluating a food safety laboratory or cosmetic testing laboratory for heavy metals work, don’t just ask which instruments they own. Ask which methods are in their ISO 17025 accreditation scope.

ISO 17025 accreditation is method-specific. A lab can hold A2LA or PJLA accreditation and still run non-accredited methods on equipment that sits next to their accredited setup. An accredited method has documented validation data — linearity, MDL, method uncertainty, matrix spike recoveries — that was independently assessed by an accreditation body assessor. That’s the documentation chain that holds up in front of an FDA investigator, an Amazon compliance review, or a plaintiff’s expert witness.

Ask for the scope of accreditation before submitting samples. It’s a publicly searchable document for any A2LA- or PJLA-accredited lab, and it shows exactly which analytes, matrices, and test methods fall under accredited status. A COA issued under an accredited method is a fundamentally different document than one issued under an in-house, unaccredited procedure — even if the instrument and the numbers look identical on paper.

The Practical Takeaway

Before specifying ICP-MS or ICP-OES on a purchase order, define the actual problem: what regulatory threshold are you defending, what’s the dominant matrix in your product, and what level of method documentation will your customer or regulator require?

For most finished dietary supplements pursuing Prop 65 or USP <232> compliance for Class 1 elements, ICP-MS is the right call for lead, arsenic, cadmium, and mercury. For mineral-dense matrices, nutritional label panels, or mid-level screening where ppb precision isn’t required, ICP-OES delivers cleaner data at lower cost. For cosmetics under MoCRA and California’s SB 312 enforcement framework, ICP-MS earns its premium for lip products and face powders. For food, ICP-OES covers most needs except the infant product category.

A testing partner worth working with doesn’t just run the machine you asked for — they push back before the samples arrive, make sure the method fits the data you actually need, and explain exactly what the accreditation scope covers. That conversation before the samples ship is often worth more than any single test result.

---

Written by Nour Abochama, Vice President of Operations, Qalitex Laboratories. Learn more about our team

Talk to our team about your testing needs. Contact us

Related from our network

• Understanding FDA Method Validation Requirements for Regulated Products — Aurora TIC covers the regulatory side of analytical method validation under 21 CFR and ICH guidelines.
• Raw Material Heavy Metal Specifications: Building a Supplier Qualification Program — Ayah Labs on setting defensible incoming material limits and COA verification for global supply chains.
• Pharmaceutical Elemental Impurity Testing Under ICH Q3D in Canada — Androxa on how Health Canada applies USP <232>/<233> and ICH Q3D to NHP and pharmaceutical products.