A standardised Mali field trial tested six consumer mosquito devices. Two heat-activated transfluthrin emanators cut catches by around 93%, while coils proved inconsistent and citronella candles and ultrasonic gadgets did nothing.
If you buy mosquito devices in any supermarket in Europe, your choices look roughly equivalent. They are not. A field study published in Frontiers in Insect Science on 21 April 2026 put six widely sold consumer products through a standardised test in Mali, and the gap between the best and the worst was not a percentage point. It was the difference between roughly 94% protection and none at all.
The study, led by Edita E. Revay and Günter C. Müller of the Malaria Research and Training Center at the University of Sciences, Techniques and Technology of Bamako, is the kind of head-to-head comparison the consumer mosquito category has needed for a long time. It also confirms something most entomologists privately suspect about a few of the products you can pick up on a shelf this weekend.
How the Test Worked
The team ran trials at two sites: a suburban neighbourhood of Bamako (Sebenikoro, Aedes- and Culex-dominated) and a Anopheles-dominated rural floodplain on the Niger. Each site held seven 20-foot test plots separated by 50-foot buffers. Inside each plot the researchers placed three COâ‚‚-baited CDC-UV traps in an equilateral triangle, four metres from the device under test. Wind speeds above 4 km/h disqualified the night; V-shaped tarp walls held the chemical plume stable when the breeze stayed honest.
Each device received eight replicates per site. Collections ran for two hours per night — 8 to 10 PM in the suburb, 11 PM to 1 AM in the field — across December 2023 and January 2024.
The six products tested:
- Thermacell® E90 — heat-activated cartridge, 5.5% transfluthrin
- Dynatrap® — heat-activated cartridge, 8.83% transfluthrin
- Hassana Mosquito Coil — 0.08% meperfluthrin
- PIC® Mosquito Repelling Coil — 0.6% pyrethrins
- Cutter® CitroGuard® Candle — 3% citronella oil
- Isotronic® Ultrasonic Repellent — oscillating frequencies, no active ingredient
That is a fair cross-section of what an average European or North American consumer would find at a hardware shop or pharmacy this summer. The point of the experiment was not to find a winner. It was to find out whether any of them deserved their shelf space.
The Hierarchy
In the suburb, the control plots caught a mean of 14.33 Aedes per trap-period. The two heat-activated pyrethroid emanators brought that to 0.88 (Thermacell) and 0.96 (Dynatrap) — reductions of 93.9% and 93.3% respectively, both statistically significant at p < 0.0001. Culex reductions in the same plots were similar: 88.7% for Thermacell, 94.3% for Dynatrap.
In the rural plot, where Anopheles gambiae catches averaged 34.96 per period, the same two devices cut numbers to 2.50 and 1.63 — reductions of 92.8% and 95.3%. Again, p < 0.0001.
The other four products produced a different picture.
The two mosquito coils (Hassana and PIC) cut Aedes in the suburb by 47.7% and 42.4% — measurable but well below the heat-activated devices, and in the rural site they actually increased Culex catches by 20% and 47%. The authors attribute that to pyrethroid coils acting more as irritants than true spatial repellents: the mosquito gets stirred up, flies further, and ends up in a CDC trap a few metres downwind. Against Anopheles in the rural site, the coils performed credibly (about 60% reduction), but the inconsistency across genera and sites means a coil is not something you can confidently rely on.
The citronella candle cut nothing to a statistically significant degree. Aedes fell 22.9% (p = 0.10); Culex and Anopheles showed no real effect. The ultrasonic device performed worse than the candle. In the suburb it actually increased Aedes catches by 7%, and in every other category its effect was indistinguishable from doing nothing at all.
This is not a marginal finding tucked into a methods section. It is the headline conclusion of the paper.
Why the Two Devices That Worked, Worked
Both Thermacell and Dynatrap use the same delivery mechanism: a small heating element that volatilises a transfluthrin-soaked mat without combustion. The active ingredient enters the air as a gas, drifts on the breeze, and reaches receptors on the mosquito at a concentration high enough to disorient or kill it. There is no smoke, no plume of partially combusted oils, no flame.
Coils and candles rely on burning. Combustion degrades a portion of the active ingredient before it ever reaches the air; the rest disperses unevenly with the smoke. The mosquito gets a noisy chemical signal — enough to move it around — but not consistently enough to protect a fixed area.
Citronella oil, even at 3%, simply does not produce enough volatile repellent at the concentrations and temperatures a candle achieves. The authors cite multiple prior evaluations reaching the same conclusion: citronella candles do not reduce mosquito landings under realistic outdoor conditions.
Ultrasonic devices have no mechanism that should work. Adult mosquitoes do not respond to the frequencies these devices emit. The category exists because the marketing predates the entomology, and survives because not enough buyers have ever measured the effect themselves. This study, again, finds none.
What This Means If You're Choosing a Device
Three things stand out.
First, the device class matters far more than the brand. Two heat-activated transfluthrin emanators from different manufacturers produced nearly identical results. The Dynatrap has more active ingredient per cartridge (8.83% vs 5.5%) but the protective effect was the same — a hint that the bottleneck is the volatilisation rate, not the dose.
Second, "natural" and "tech-forward" do not map to effective. The citronella candle scores well on green credentials and consumer appeal; it does almost nothing in a controlled field trial. The ultrasonic device looks futuristic; it does less than nothing.
Third, every device on this list relies on insecticidal chemistry to work at all — and every device that works relies on pyrethroids. That is a strategic issue, not a moral one. The authors flag it explicitly: as pyrethroid resistance grows in field populations, the protective effect of these emanators will erode. Pyrethroid-based vector control is running on a budget that is already being drawn down by Anopheles funestus nAChR mutations now turning up in 13 African countries.
What Mosticare Takes From This
Two careful conclusions.
The first is honest. If you are using a consumer device to keep mosquitoes off a terrace at dusk in a high-pressure environment, a heat-activated transfluthrin emanator is the only device class with reproducible field evidence that it works. A citronella candle is a piece of dinner-party staging. An ultrasonic gadget is a placebo.
The second is structural. Chemical emanators are buying time, not solving the problem. They depend on a single class of insecticide; that class is losing ground to resistance; and the more we lean on the emanators, the faster that erosion happens. The honest long-term answer for a family that wants to be outside after dark — or to sleep with the windows open — is the physical barrier that does not depend on chemistry: the screen, the net, the curtain. The mosquito that cannot reach the skin does not need to be repelled.
That is not a story the device industry tells well. The shelf space pays for the chemistry. Studies like this one are the rare moment when someone with no incentive to sell anything actually measures what is happening, and lets the rest of us read the result.
What we know
- In a Mali field trial, two heat-activated transfluthrin emanators (Thermacell E90, Dynatrap) reduced Aedes catches by 93.3–93.9% and Anopheles catches by 92.8–95.3% versus untreated controls (p < 0.0001). (Revay et al., Frontiers in Insect Science, 2026)
- Two mosquito coils tested (Hassana 0.08% meperfluthrin; PIC 0.6% pyrethrins) produced inconsistent results — moderate against Aedes in the suburb (~42–48%), but actually increased Culex catches in the rural site by 20–47%. (Revay et al., 2026)
- A citronella candle (Cutter CitroGuard, 3%) and an ultrasonic device (Isotronic) produced no statistically significant protection in any genus at any site. (Revay et al., 2026)
- The WHO 2025 Guidelines for Malaria formally recognise spatial repellents as a recommended intervention class — making field-based device evaluation directly relevant to public-health policy. (WHO, 2025)
- The authors caution that pyrethroid resistance threatens long-term effectiveness of all emanator-class products and call for new modes of action. (Revay et al., 2026)
Sources Cited
- Revay EE, McKenzie K, Junnila A, Styer K, Prozorov AM, Traore MM, Cui L, Yakovlev RV, Saldaitis A, Traore SF, Beavogui AH, Prozorova TA, Petrányi G, Benz U, Xue R-D, Müller GC. "Performance of different spatial repellents (spatial emanators) against vector mosquito species in Mali, West Africa: a field trial using a non-human test method." Frontiers in Insect Science 6:1811511, 21 April 2026. DOI: 10.3389/finsc.2026.1811511. https://www.frontiersin.org/journals/insect-science/articles/10.3389/finsc.2026.1811511/full
- World Health Organization. Guidelines for Malaria. 2025: https://www.who.int/publications/i/item/guidelines-for-malaria