Skip to main content
Back to Lower-Contact Product Guides

Are Candles Toxic? Soy, Paraffin and Beeswax Compared

Last reviewed: by the MicroPlastics Research Desk. Submit a correction or see our editorial standards.

Quick answer

Every candle releases fine particles and gases as it burns, whatever the wax. In a small 2026 study of six scented candles in a closed room, the soy and beeswax candles gave off far more ultrafine particles, while the paraffin candles gave off more of the larger 0.3 to 10 µm particles. So “soy is non-toxic” is too simple. Lead-cored wicks have been effectively banned in the US since 2003. A trimmed wick, no drafts and fresh air are the habits with support behind them. The table below sets out what the study found for each wax.

Check the containers in your kitchen

Scan what you heat and store food in to see what drives each score, and one swap worth making.

Download on the App StoreGet it on Google Play

Free · 3 scans · iOS and Android

Are Candles Toxic? Soy, Paraffin and Beeswax Compared

Key Takeaways

  • No wax burns clean. All six candles in the 2026 study raised particle levels well above the room’s background.
  • “Natural” candles made more of the smallest particles. Soy and beeswax produced about 12 times more ultrafine particles (10 to 300 nm) on average than paraffin.
  • Paraffin made more of the larger particles. The paraffin candles led for particles from 0.3 to 10 µm.
  • The sample was tiny. The study tested three paraffin, two soy and one beeswax candle, burned for one hour each in an unventilated kitchen. It does not rank brands.
  • Drafts make candles sootier. A Lund University chamber study found that flickering, “stressed” burning released mostly soot, and that preventing it can reduce exposure.
  • Lead wicks are capped by law. Since October 15, 2003, US candle wicks with a metal core may contain no more than 0.06% lead by weight.

Candles and indoor air, by the numbers

mean ultrafine particle count near soy and beeswax candles
8.4 × 10¹¹ /m³
vs 7.0 × 10¹⁰ /m³ for paraffin; 10 to 300 nm, closed 37 m³ room, background subtracted
Crunkelton et al., Atmosphere 2026
mean 0.3 to 10 µm particle count near paraffin candles
2.0 × 10⁷ /m³
vs 2.8 × 10⁶ /m³ for soy and beeswax; same room and conditions
Crunkelton et al., Atmosphere 2026
maximum lead in a US candle wick’s metal core
0.06%
by weight, for candles made or imported from October 15, 2003
16 CFR 1500.17(a)(13)
US home fires a year started by candles
5,894
annual average, 2020 to 2024; 63 deaths a year
NFPA

What does a burning candle release?

A flame burns wax vapor, and that is never perfectly complete. The EPA lists burning candles among indoor sources of particulate matter, alongside cooking, fireplaces and space heaters. It notes that particles smaller than 10 micrometers can get deep into the lungs. Candles also release gases, and scented candles add fragrance compounds to the mix.

What matters is particle size. Researchers often split candle particles into ultrafine particles, smaller than about 0.1 µm (100 nm), and larger fine particles measured in the 0.3 to 10 µm range. Ultrafine particles carry little mass but are very numerous and reach the deepest parts of the lung. That is why a wax that does well on one size can do worse on another.

Soy vs paraffin vs beeswax: what the 2026 study found

The study behind the recent “natural candles aren’t cleaner” headlines is Crunkelton and colleagues, from Georgia Southern University, published in Atmosphere on May 18, 2026. They bought six candles that were mostly scented: three paraffin candles labelled synthetic, and three labelled natural (two soy and one beeswax). Each was burned once for one hour in a closed, unventilated kitchen of about 37 m³. Instruments at 5 cm, 1 m and 2 m from the flame, at breathing height, took readings at 15, 35 and 55 minutes, and the room’s background level was subtracted.

Mean results by candle group in Crunkelton et al. 2026 (closed 37 m³ room, 1-hour burn, background subtracted)
MeasureSoy + beeswax (n = 3)Paraffin (n = 3)What it means
Ultrafine particles, 10–300 nm8.4 × 10¹¹ ± 4.1 × 10¹¹ per m³7.0 × 10¹⁰ ± 6.6 × 10¹⁰ per m³Natural candles higher (p < 0.001). Soy and beeswax did not differ from each other.
Lung-deposited surface area (LDSA)195.2 ± 167.0 µm²/cm³82.1 ± 77.9 µm²/cm³Natural candles higher (p = 0.002). The single beeswax candle was higher than the soy candles.
Particles 0.3–10 µm2.8 × 10⁶ per m³2.0 × 10⁷ per m³Paraffin higher (p < 0.001). The authors link this to soot.
Total VOCs669 ± 276 ng/L442 ± 134 ng/LNatural higher on average, but not statistically significant.

The authors conclude that their results “challenge common perceptions that natural candles represent inherently cleaner alternatives.” They suggest that plant-based waxes and multi-wick designs may burn less completely, favoring ultrafine particles, and that paraffin is the stronger source of larger, soot-type particles. Toluene and styrene showed up only at low levels in both groups.

The limits matter as much as the headline:

  • Six candles. There was one beeswax candle, so “beeswax” here means a single product. The candles also differed in size and wick count, and one soy candle had a wooden wick.
  • A worst case. The room had no ventilation by design. The authors say this limits how well the results apply to ordinary homes.
  • Scented candles. The authors say they could not evaluate unscented products, so the results mix wax effects with fragrance effects.
  • Room conditions. Door openings and people in the room may have added variability, and averaging over time may have hidden short peaks.
  • One reporting mismatch. The paper’s text describes paraffin as producing “nearly double” the larger particles in each size range, while its reported means differ about sevenfold. We show the means as published.

The study reported no external funding and no conflicts of interest. It measured air concentrations, not health outcomes in people.

What earlier candle studies show

Two 2021 chamber studies from Lund and Aarhus universities, published in Indoor Air, help explain why the same candle can behave very differently.

  • Drafts change everything. Andersen and colleagues burned five types of pillar candle under fluctuating air currents in a 21.6 m³ chamber. Under this “stressed” burning, particle mass was dominated by soot (black carbon). Wax and wick composition strongly affected soot, PM2.5 and particle-bound PAHs, but not NOx or formaldehyde. Their conclusion: preventing stressed burning “can reduce exposure to pollutants in indoor air.”
  • The first hour is dirtier. Rasmussen and colleagues burned five candle types steadily for six hours. For most candles, particle number, size and mass were higher in the first hour than in the stable phase that followed. The authors suspect wick additives released at ignition.

Together these point to the same practical lesson as the 2026 study. How you burn a candle, and what its wick is made of, can matter as much as the wax on the label.

Wicks: lead, zinc, cotton and wood

Lead-cored wicks are the one candle hazard with a federal rule. In 2003 the Consumer Product Safety Commission banned lead-cored wicks and candles made with them, after finding that some tested candles emitted more than 3,000 micrograms of lead per hour. The rule, at 16 CFR 1500.17(a)(13), applies to candles and metal-cored wicks made or imported on or after October 15, 2003. It caps lead at 0.06% of the metal core’s weight. CPSC says metal-cored wicks today are most likely zinc, and that most makers use zinc, cotton, paper or synthetic fibers. You cannot tell lead from zinc by looking. For an old or unbranded candle, CPSC and Health Canada both suggest asking the seller.

Wooden wicks are popular, but we found no peer-reviewed emission data that isolates them. The one wooden-wick candle in the 2026 study was a soy candle, so its results cannot be separated from the wax.

Do candles release microplastics?

We found no study measuring microplastics from burning candles. The plastic questions are about decoration and packaging. Glitter candles are the obvious one: conventional glitter is plastic film, as our glitter guide explains, so glitter on or in a candle is worth skipping. Plastic shrink wrap and lids are ordinary packaging waste. For the plastic particles that actually fill indoor air, mostly fibers from carpets and clothing, see microplastics in the air at home.

Keep going · Chapter 10 of 10

A 30-day plan and shopping lists

Four weeks of changes in order of cost, shopping lists by price band, and the five things to do if you only do five.

  • One chapter of Less Plastic, Room by Room, 77 pages
  • 22 swaps, each with a cost band, and a 30-day plan
  • 165 studies cited, to the same standard as this article
Get the guide · $19One payment · instant PDF download

Use the App

Cut the plastic sources you can control

Scan home and gift products to see their materials, glitter and plastic packaging included, with lower-plastic alternatives.

Download on the App StoreGet it on Google Play

How to burn candles with less smoke

Each habit below is tied to the source that supports it. None has been tested as a way to prevent a health outcome.

  1. Trim the wick to about 1/4 inch (5 to 7 mm) before lighting, and again every 2 to 3 hours to prevent high flames. That is Health Canada’s advice.
  2. Keep candles out of drafts. Away from open windows, fans and vents. Flickering, stressed burning produced mostly soot in the Lund chamber study.
  3. Air the room after burning. The 2026 study chose a room with no ventilation at all as its worst case. Ventilate after the flame is out, so that a fan or open window does not blow on the flame itself.
  4. Burn fewer candles at once, and fewer, longer sessions rather than many short ones, since the first hour emitted more particles in the steady-burn study. This is our inference from that study, not a tested rule.
  5. Skip candles you cannot identify, such as very old or unbranded metal-wick candles, or ask the seller what the wick core is.
  6. Keep the flame at least 12 inches from anything that can burn, and blow it out before you leave the room or go to sleep. That is NFPA advice. Fire is the best-documented candle harm: about 5,894 US home fires a year, with Christmas the peak day.

If you use an air purifier, our guide to whether air purifiers remove microplastics explains how HEPA filters and room size interact.

What about flameless candles and diffusers?

A battery LED candle has no flame, so it adds no combustion particles and no fire risk. The trade-off is a plastic body and batteries. Reed diffusers and wax melts avoid the flame, but scented ones are made to release fragrance into the room. We did not find comparable emission data for these alternatives, so we can describe them only by what they lack: combustion.

So, are candles toxic?

“Toxic” depends on dose, and no study we found ties ordinary candle use to a specific illness. What the evidence does support: every candle adds particles to indoor air, “natural” waxes are not automatically cleaner, drafts and the first hour of burning make things worse, and lead wicks have been capped by US law since 2003. If you enjoy candles, burning fewer of them, in still air, with trimmed wicks and a room you air out afterwards is a reasonable middle path. If you are giving one as a gift, our low-plastic gift guide has other ideas.

What the evidence does not establish

  • A ranking of candle brands. The 2026 study tested six unnamed products and does not identify brands.
  • That beeswax is cleanest or dirtiest. One beeswax candle cannot represent the category.
  • Exposure in a normal, ventilated home. The 2026 room was deliberately unventilated.
  • Health effects from a candle habit. These studies measured air, not people.
  • Emissions from wooden or coconut-wax wicks and candles on their own. We found no study that isolated them.

What the MicroPlastics app checks

  • Materials and packaging a maker discloses for scanned home and gift products, including plastic glitter and wraps.
  • Lower-plastic alternatives in the same category.
  • Product information only: the app does not measure candle smoke, particles or indoor air.

Frequently Asked Questions

Are soy candles non-toxic?

No candle is emission-free. In a 2026 study of six scented candles in a closed room, the soy and beeswax candles produced about 12 times more ultrafine particles on average than paraffin, though paraffin produced more larger particles. Soy is not automatically cleaner.

Is paraffin or soy better for indoor air?

It depends on particle size. In the 2026 Atmosphere study, paraffin candles produced more 0.3 to 10 µm particles, while soy and beeswax produced more ultrafine particles and higher lung-deposited surface area. The sample was only six candles in an unventilated room.

Are scented candles bad for your lungs?

All candles add particles to indoor air, and the EPA lists candles among indoor particle sources. In the 2026 study, all the candles were scented, so wax and fragrance effects could not be separated. Studies we found measured air, not lung health in people.

Are beeswax candles the safest?

The evidence does not show that. The 2026 study included one beeswax candle, and it had higher lung-deposited surface area than the soy candles. One product cannot represent all beeswax candles.

Do candles still have lead wicks?

Not legally in the US. Since October 15, 2003, candles with metal-cored wicks may contain no more than 0.06% lead in the metal core. CPSC says metal cores today are most likely zinc. You cannot tell by looking, so ask the seller about old or unbranded candles.

Do candles release microplastics?

We found no study measuring microplastics from candle burning. Glitter on candles is usually plastic, and wraps and lids are plastic packaging, so those are the parts to avoid if plastic is your concern.

How can I make candles produce less smoke?

Trim the wick to about 1/4 inch before lighting (Health Canada), keep the flame out of drafts (stressed burning was mostly soot in a Lund study), burn fewer candles, and air the room afterwards.

Are candles safe around kids and pets?

The best-documented risk is fire: NFPA counts about 5,894 US home candle fires a year. Keep flames 12 inches from anything that can burn and out of reach, and blow them out before leaving the room. Flameless candles avoid both fire and combustion particles.

Sources

  1. Crunkelton D, Ilie M, Seybold D, Soo J-C, Adhikari A (2026). Emission Characterization of Synthetic and Natural Candles in a Residential Environment. Atmosphere 17(5):515. doi:10.3390/atmos17050515.
  2. Andersen C, Omelekhina Y, Rasmussen BB, et al. (2021). Emissions of soot, PAHs, ultrafine particles, NOx, and other health relevant compounds from stressed burning of candles in indoor air. Indoor Air 31(6):2033–2048. doi:10.1111/ina.12909.
  3. Rasmussen BB, Wang K, Karstoft JG, et al. (2021). Emissions of ultrafine particles from five types of candles during steady burn conditions. Indoor Air 31(4):1084–1094. doi:10.1111/ina.12800.
  4. US Consumer Product Safety Commission (Retrieved November 22, 2026). 16 CFR 1500.17(a)(13): Candles with metal-cored wicks and metal-cored candlewicks. eCFR.
  5. US Consumer Product Safety Commission (2003). CPSC Bans Candles With Lead-Cored Wicks (Release 03-105, April 7, 2003). cpsc.gov.
  6. Health Canada (Retrieved November 22, 2026). Candle Safety. canada.ca.
  7. US Environmental Protection Agency (Retrieved November 22, 2026). Indoor Particulate Matter. epa.gov.
  8. National Fire Protection Association (Retrieved November 22, 2026). Candle Fire Statistics and Safety (2020–2024 data). nfpa.org.

Compare your kitchen contact pattern over time

The app turns saved scans into a modeled household contact pattern based on what you cooked in, stored in, and heated this month. The trend describes the assessed mix, not health or bodily exposure.

Download on the App StoreGet it on Google Play
  • iOS and Android
  • Current terms in each store

“Really cool to scan stuff around the kitchen and see what's actually in it. The swaps it suggests are realistic.” App Store review · 5.0★

Related Research