VapeRisk Lab acquired 22 disposable devices in March 2026 from regular U.S. retail (gas station, convenience, and online sources — no manufacturer samples). Aerosol from each device was collected on standard quartz filters under ISO 20768 puffing parameters and analyzed by ICP-MS at Eurofins' Madison, WI lab. The headline finding: only 4 of 22 devices cleared the FDA's draft thresholds for all four heavy metals tested.
Methodology
For full methodology, see Section 5 of the PDF. All raw data is published alongside this report and is available for re-analysis under our open-data policy.
Results — by element
Full results — all 22 devices
| Device | Pbµg/m³ | Crµg/m³ | Niµg/m³ | Cdµg/m³ | Verdict |
|---|---|---|---|---|---|
| Lost Mary MO20000Disposable · 20K | 0.9 | 2.1 | 3.4 | <0.1 | Pass |
| Elf Bar BC25000Disposable · 25K | 1.4 | 3.0 | 4.1 | <0.1 | Pass |
| RAZ TN9000 CrystalDisposable · 9K | 0.6 | 2.8 | 2.9 | <0.1 | Pass |
| Funky Republic Ti7000Disposable · 7K | 2.2 | 4.5 | 5.8 | 0.2 | Pass |
| Geek Bar Pulse 25KDisposable · 25K | 5.8 | 9.4 | 7.1 | 0.3 | 3 of 4 |
| HQD Cuvie SlimDisposable · 6K | 5.2 | 5.6 | 11.8 | 0.4 | 2 of 4 |
| Breeze ProDisposable · 6K | 8.4 | 12.1 | 14.3 | 0.6 | Fail |
| Kang Vape ONEE ProDisposable · 12K | 6.4 | 10.8 | 7.6 | 0.9 | Fail |
| Hyde IQ RechargeDisposable · 5K | 19.4 | 22.6 | 15.4 | 1.1 | Fail |
| Esco Bars MegaDisposable · 5K | 47.2 | 14.8 | 28.7 | 2.4 | Fail |
| FlumPebble ProDisposable · 6K | 7.1 | 6.4 | 12.0 | 0.5 | Fail |
| Spaceman Prism 20KDisposable · 20K | 3.8 | 8.9 | 5.4 | 0.4 | 3 of 4 |
| Vozol Gear 10000Disposable · 10K | 9.8 | 11.6 | 13.2 | 1.4 | Fail |
| Air Bar BoxDisposable · 5K | 8.7 | 10.4 | 12.6 | 0.8 | Fail |
Discussion
The dispersion in our results — a 78× spread on lead alone — is consistent with the hypothesis that heavy-metal contamination in disposables is driven primarily by manufacturing-side variables (alloy choice, coating quality, atomizer geometry) rather than by user behavior or storage. Devices from manufacturers with public quality-control documentation cluster strongly on the safer side; devices with no published QC data show the widest variance.
This is the same pattern Olmedo et al. (2024) identified in refillables, and the same mechanism the FDA cited in its draft guidance last September. Wednesday's PMTA enforcement action targeted four of the eleven devices that failed our screen.
What this means for buyers
For buyers: of the 22 devices we tested, four passed all four thresholds and three more passed three of four. If you're buying disposable, prefer the four "Pass" devices in the table above until further data is available. If your local retailer doesn't carry any of them, that itself is a signal worth taking seriously.
VapeRisk Lab will re-test the failing devices in our August 2026 cycle. We will also expand the screen to include carbonyls (formaldehyde, acetaldehyde, acrolein) — the parallel safety axis we know less about for high-puff disposables.
This study was funded entirely by VapeRisk operating revenue (subscription + non-affiliate ad). No manufacturer provided samples, funding, or input on methodology or results. All 22 devices were purchased anonymously at retail. The lead author and co-authors declare no relevant financial conflicts. Raw data, instrument calibration logs, and chain-of-custody records are available on request to peer reviewers.