Biological Pest Control: What Is Actually Registered
The EPA has registered 457 biopesticide active ingredients, by our count of its own list, up from 70 in 1990. EPA publishes no total itself and has not refreshed its product count in roughly a decade. In structural pest control the field narrows to three things that work: Bti for mosquito larvae, a Beauveria bassiana product for bed bugs, and beneficial nematodes in soil. Classical biological control, the kind that establishes and keeps working, essentially does not exist for structural pests.
Biological control carries a lot of expectation: release a natural enemy, it establishes, the problem manages itself. That is a real thing in agriculture and forestry. Inside buildings it is close to fiction, and the honest version of the story is more useful than the hopeful one.
What is actually registered
The EPA defines a biopesticide in three categories: microbial, plant-incorporated protectants, and biochemical. Counting the rows of the agency’s own biopesticide active ingredient list gives 457 actives as of its last update in July 2026, against 70 registered by 1990.
Three caveats, because the number deserves them. EPA publishes no total of its own, so this is our count of their list. Their wording is “have been registered”, which is cumulative-ever and may include cancelled actives. And the page currently carries an unfilled placeholder reading “as of July XX, 2026”, so we have used the last-updated stamp instead.
There is no current product count at all. The agency stopped publishing one. The historical figures it does still carry contradict each other: one archived page states 436 active ingredients and 1,401 products as of September 2015, while a 2021 snapshot states 299 active ingredients and the identical 1,401 products as of April 2016. A 137-ingredient drop in seven months alongside an unchanged product total is not a real-world event, it is a documentation problem.
Recent growth runs at roughly 4 to 17 new active ingredients a year. The direction is clear even where the bookkeeping is not.
The three that work in structural pest control
Bti, Bacillus thuringiensis israelensis, is the strongest case in the whole field. Five registered strains across 48 products. Field trials report 100 per cent larval mortality within 24 hours, and emergence inhibition of 80 per cent or better out to 21 days under the WHO protocol. One high-rise Aedes programme recorded reductions of 87.4, 93.8 and 98.5 per cent.
The decisive argument for Bti is what has not happened: no documented field resistance after decades of use, which cannot be said of any conventional insecticide class in this industry. Its limits are honest ones. It is a larvicide only, it degrades in sunlight, and it has to be eaten. It is the backbone of the public programmes covered in our guide to mosquito control districts.
Beauveria bassiana is registered as a bed bug product. Published trials report 98 to 100 per cent mortality on susceptible strains and over 94 per cent on three pyrethroid-resistant strains, with median survival around four to five days and transfer between exposed and unexposed bugs. The label claims three months of residual.
Two limits matter. It takes four to ten days to kill, which is a different operational proposition from a knockdown, and no independent apartment or field trial has been published. The performance evidence is the registrant’s and the developing university’s.
Beneficial nematodes genuinely work in soil. Steinernema scarabaei has given 75 to 95 per cent control of third-instar Japanese beetle, oriental beetle and European chafer grubs, and one combination outperformed chlorpyrifos. They fail outside a narrow envelope: soil above 55 and below 90 degrees Fahrenheit, 20 to 25 per cent soil moisture, pre-irrigation, no application in the middle of the day, uncompacted soil, small larvae and live product. Most failures are application failures, not product failures.
The canonical failure, worth knowing
Metarhizium is the fungus that should have worked on termites and did not. A product was registered in the United States in 1994 and withdrawn for lack of virulence and low field efficacy.
The reason is the interesting part. Termites groom each other, detect diseased nestmates, and wall them off. Nest carbon dioxide and naphthalene inhibit fungal growth. And virulence and repellency are partly inversely linked, so the more lethal the isolate, the better the colony avoids it. A social insect with hygienic behaviour is a hard target for a pathogen you apply from outside.
Against bed bugs the same genus performs well only at around 98 per cent relative humidity and poorly at ambient, and one assessment called it a poor pathogen for use in control. There is no EPA-registered fungal termiticide today, which is worth remembering whenever a product claims otherwise. Our guides to termite treatment and termite bonds cover what is actually used.
Why none of this is classical biological control
Classical biological control means introducing a natural enemy that establishes and keeps suppressing the pest without further input. Everything structural pest control actually uses is augmentative or inundative: a pesticide whose active ingredient happens to be alive, applied, reapplied, and never establishing.
The best-documented attempt at the real thing is a parasitic wasp, Aprostocetus hagenowii, released against American cockroaches. It achieved 18.1 per cent parasitism of egg cases in wall crevices and 13.3 per cent in sewers, and required 189 and 428 released females respectively per egg case parasitised. That is not control, and it is not economic.
There is a cautionary tale on the other side too. The Asian lady beetle was introduced deliberately as a biological control agent for aphids and is now one of the more complained-about structural nuisance pests in the country, as our guide to Asian lady beetles sets out. Biological agents that do establish are not reversible.
What this means in practice
- Judge a biological product as a pesticide, because that is what it is: it has a label, a rate, an application window and a reapplication interval.
- Expect slow. Four to ten days for the bed bug fungus, days for nematodes. None of these produce a knockdown, which is the same expectation problem that dogs neem oil.
- Respect the envelope. Nematodes fail on temperature, moisture and timing far more often than on efficacy.
- Treat “natural” as a claim about the active ingredient, not about safety or effectiveness, which is the theme of our review of natural pest control products.
- Be sceptical of fungal termite claims. One was registered and withdrawn, and nothing has replaced it.
One development worth watching: the EPA list now includes two registered Wolbachia products, incompatible male mosquitoes for Aedes aegypti and Culex quinquefasciatus, registered in 2024 and 2026. A registered pesticide whose product is a living male mosquito is a genuinely new category. We have not found US efficacy data for either yet.
How we did this
The registration figures are our own count of the rows in the EPA’s published biopesticide active ingredient list, retrieved in October 2026, with the cumulative time series built from that list’s year-first-registered column. We have said plainly where EPA’s own numbers conflict rather than picking the convenient one.
Efficacy figures come from published field and laboratory trials. Where the only evidence is the registrant’s own, as with the bed bug fungus, we have said so. We did not obtain a plant-incorporated protectant count, and we found no US efficacy data for the Wolbachia registrations.
Frequently asked questions
What is a biopesticide?
An EPA regulatory category covering pesticides derived from natural materials, in three groups: microbial, plant-incorporated protectants and biochemical. By our count of EPA’s own list there are 457 registered biopesticide active ingredients as of July 2026, up from 70 in 1990. EPA itself publishes no total and has not updated its product count in about a decade.
Does biological pest control work indoors?
In three specific applications. Bti gives near-total mosquito larval mortality with no documented field resistance after decades. A registered Beauveria bassiana product kills bed bugs, including pyrethroid-resistant strains, over four to ten days. Beneficial nematodes control soil grubs within a narrow temperature and moisture envelope. Outside those, the evidence is thin.
Is there a biological treatment for termites?
No registered one in the United States. A Metarhizium fungal product was registered in 1994 and withdrawn for lack of virulence and low field efficacy, because termites groom each other, wall off diseased nestmates, and their nest atmosphere inhibits fungal growth. More virulent isolates tend to be more strongly avoided.
Is biological control safer than conventional pesticides?
Often, but it is not automatic, and the agents are regulated as pesticides with labels and restrictions for that reason. Bti has an unusually clean non-target record. By contrast the Asian lady beetle was introduced deliberately as a biological control agent and is now a significant structural nuisance pest and a documented indoor allergen, which is what an irreversible introduction looks like.
Sources
- US Environmental Protection Agency, Biopesticide Active Ingredients list, retrieved October 2026, and archived biopesticide registration pages.
- Barbarin and colleagues on Beauveria bassiana against bed bugs, and Ulrich and colleagues on Metarhizium.
- Published Bti field trials and WHO protocol emergence-inhibition studies.
- Koehler and colleagues, Aprostocetus hagenowii parasitism of American cockroach egg cases, Journal of Economic Entomology 104:2031, 2011.
- Published entomopathogenic nematode trials against white grubs.
