Heat or Chemical for Bed Bugs? What the Research Says

Halo detects and reports. We do not treat, we do not sell heat equipment, and we do not sell insecticide. That means we have nothing riding on the answer to this question, which is unusual in this industry and is the only reason this article is worth reading.

Search this question and you will get a page of results. A heat equipment manufacturer explaining why heat wins. Five pest control companies, each of which explaining why the method they own is better. The conclusion in every one of those articles matches the invoice the company would like to send you. That is not a conspiracy, it is just what happens when everyone answering a question has a stake in the answer.

So this is what the published entomology says, including the parts that are inconvenient for both camps.

Start with the thing that should be said first and almost never is. There is no peer reviewed head to head field trial of whole room heat against a chemical program with elimination as the endpoint. It does not exist. The comparative literature is chemical against chemical, chemical against integrated pest management, and steam against spray. Whole room heat has essentially never been placed in a controlled arm against a chemical program.

That matters because the entire argument you will hear on the phone is conducted as though the question has been settled. It has not been studied. Any company telling you their method is proven superior is describing a study that has never been published.

What does exist is a large body of field research on what each approach achieves on its own, and that research is more useful than the sales pitch.

The field results are worse than most people expect. In a controlled comparison in a low income apartment building published in Insects in 2012, three strategies were run side by side over ten weeks. Insecticide only eliminated bed bugs in 33 per cent of units. Non chemical methods, meaning steam, vacuuming and mattress encasements, eliminated them in 67 per cent. The combined integrated approach achieved a 99.9 per cent reduction in counts.

One caveat on that study, because it gets quoted without it. The insecticide group started with roughly five times the bed bug counts of the non chemical group. The comparison is not clean. But it is the closest thing to a controlled three arm trial anybody has run, and the direction is consistent with everything around it.

The broader picture from a decade of field studies reviewed in the Journal of Integrated Pest Management in 2017 is that chemical programs typically achieve 75 to 85 per cent reduction, while integrated programs combining chemical and non chemical methods reach 90 to 96 per cent. Individual studies keep finding the same residue. Five to eight per cent of infestations persisted through treatment in one. Live bugs remained in three of twenty two apartments at twelve weeks in another. In a 2014 study using three different products, all achieved over 90 per cent count reduction, but only 22 per cent of apartments were actually clear at twelve weeks.

The review's own conclusion is worth quoting because it is blunter than anything a treatment company will tell you. Complete eradication in most places continues to be problematic, especially when the control action is assessed as a terminal effort.

This is the best documented material in the entire field and it is not close.

Pyrethroids are the class most commonly used against bed bugs. A strain studied at Virginia Tech and published in PLOS ONE in 2011 showed roughly 5,200 fold resistance to deltamethrin, through both target site mutation and metabolic detoxification. An Australian strain measured in 2015 recorded resistance factors of approximately 370,000 for deltamethrin and 1,235,000 for permethrin, with under 60 per cent mortality even at the maximum dose tested.

The genetic marker for this is widespread rather than local. Work published in 2018 found the L925I mutation in 93.5 per cent of human associated European bed bug populations. A 2015 study of Israeli populations found ten of twelve at 100 per cent frequency for the same mutation.

The neonicotinoids were introduced as the answer to pyrethroid resistance. They are not holding either. Research published in the Journal of Medical Entomology in 2016 measured resistance ratios against a susceptible reference strain and found acetamiprid resistance ratios exceeding 33,000 in two of three field populations, thiamethoxam at 546, and imidacloprid at 463.

It has now started appearing in the products marketed as alternatives. Reduced susceptibility to chlorfenapyr was documented in three field strains in 2017. A pyrethroid resistant strain survived 50 per cent longer than a susceptible one on silica desiccant at sub label rates. Eggs and first instar nymphs are dramatically harder to kill than adults, with lethal concentrations 39 to over 1,900 times higher in one study.

One honest gap. There is no published Canadian or Ontario resistance survey. All of the North American resistance data is American. The Canadian anchor is a 2023 CMAJ article stating that many bed bug populations are resistant to pyrethroids, without Canadian numbers behind it. Nobody has measured what is in Toronto buildings specifically.

Heat has one enormous advantage and it is not the one usually advertised. Bed bugs do not appear able to evolve resistance to it.

Researchers at Purdue ran seven generations of heat selection on a bed bug colony and published the result in PLOS ONE in 2019. Survivorship rose modestly in the second and third generations to around 50 to 55 per cent, then stopped. It did not increase past the fourth generation, and the colony failed to produce an eighth. Compare that to the 5,200 fold and 370,000 fold numbers above and you have the strongest single argument for heat in the literature.

That same study is where the commonly quoted lethal figures come from. At 45 degrees, median lethal time was 14.3 minutes and 99 per cent mortality came at 23.2 minutes.

Those numbers need a caveat, and it is a large one that heat companies rarely supply. A separate study published in Insects in 2011 measured 94.8 minutes for adults at that same 45 degrees, and just over seven hours for eggs. That is a sixfold discrepancy with the Purdue figure at an identical temperature, and no published work reconciles them. Real field protocols are built on the conservative number, which is why a genuine heat treatment runs for hours rather than twenty minutes. Eggs are the binding constraint, needing roughly 54.8 degrees for 99 per cent mortality.

The best field data on this comes from Virginia Tech, presented at the International Conference on Urban Pests in 2017. Researchers instrumented fourteen occupied apartments during commercial heat treatments and recorded how long each location took to reach lethal temperature.

Near the heater output, about an hour. Between couch cushions, an hour and three quarters. Inside dresser drawers containing clothes, about three hours. The exposed floor and wall junction, three hours and eleven minutes. A utility closet floor, three and a half hours. Under piles of clothing, about four hours.

Survivors were documented. Live adults in cold spots, and eggs that went on to hatch in three separate units. In one apartment the sensor at the egg location read only 48 degrees when the treatment was terminated, below the egg threshold.

The finding that surprises people: clutter made no statistically significant difference. Low, medium and high clutter units averaged two hours fifty one, three hours four, and three hours thirty two minutes. The authors concluded that cold spots, not clutter, were the single greatest challenge to thermal remediation.

Virginia Cooperative Extension puts the same point more directly in its guidance for homeowners. Ambient temperature, meaning the temperature of the heated air in the home, is not relevant. What matters is the temperature at the harbourage, which is exactly the measurement a company can only make if it has placed sensors there and watched them.

Heat also leaves nothing behind. The Virginia Department of Agriculture states plainly that heat has no residual activity and that bed bugs can re-infest the day after treatment. In a Toronto apartment building with an untreated neighbour, that is not a theoretical risk.

Every Canadian figure available is from an industry source. There is no published cost effectiveness analysis comparing the two methods in any market, let alone this one. Treat these as market pricing, not as research.

Ontario pest control companies publishing 2026 price guidance put chemical treatment at roughly $300 to $900 for a room or small apartment, usually requiring two to four visits spaced over three to six weeks. Heat treatment runs roughly $1,000 to $3,000, typically in a single visit. Combined heat with a chemical perimeter sits around $1,200 to $4,000.

The price gap is largely capital cost rather than labour. Heat equipment runs from about $11,000 for a small electric unit to over $150,000 for a trailer mounted system. A sprayer costs a fraction of that. That is the structural reason heat is expensive, and it is worth knowing when a company explains the price to you.

For scale, the only per unit cost figure in the peer reviewed literature is from a 2009 study measuring two integrated pest management programs in apartments at roughly $463 and $482 US per unit over ten weeks, achieving 50 per cent eradication in both at that point.

This is the part that ends the argument, and it is not close either. Virginia Cooperative Extension, the Virginia Department of Agriculture, Purdue entomology, the Canadian Medical Association Journal and Clinical Microbiology Reviews all recommend the same thing, which is combining methods rather than choosing one.

Virginia Cooperative Extension specifically advises applying a residual silica aerogel after a heat treatment completes, to the floor and wall junctions, drywall voids, closet floors and electrical outlets. That is heat for the kill and chemical for the cold spots and the re-entry, which addresses both documented failure modes at once.

There is direct experimental support for combining. A study in the Journal of Pest Science in 2017 found that desiccant dust plus carbon dioxide eradicated all five infested dormitories tested, while desiccant dust alone failed in six. Another found dichlorvos strips alone took seven days to reach complete mortality, strips with a fan took three days, and strips with a fan and heat took 36 hours.

The reason is structural rather than chemical. Heat and steam penetrate crevices that a liquid spray cannot reach. Spray leaves a residual that heat cannot provide. Their weaknesses do not overlap, which is the textbook condition for combining two tools.

So the honest answer to which is better is that the framing is wrong. The research community does not treat this as an open question, and the only people who present it as a binary choice are the people selling one of the two options.

Across all of this research, one variable predicts results more reliably than the treatment method, and it is whether anyone knows where the bed bugs are and whether anyone checks afterward.

A study of a building wide program in affordable housing found 15 per cent of 358 apartments infested at baseline, and management was unaware of 71 per cent of them. Proactive inspection found 71 per cent of all infestations identified over the study period. Complaint driven reporting would have missed roughly seven in ten.

Over 24 months across three low income communities, buildings running proactive inspection programs reduced infestation rates by 49 and 64 per cent. The building running conventional complaint driven pest control managed 26 per cent, and had a new introduction rate of 11 per cent against 3 and 7 per cent at the proactive sites.

The detection method matters too. In a study of 2,372 apartments, interceptor monitors left in place detected 89 per cent of infestations and a brief visual inspection detected 72 per cent. Combined, only two of 291 infestations escaped both.

None of that is an argument for a particular treatment. It is an argument that the treatment decision is downstream of a detection decision most people never make.

There is no validated protocol for declaring a home bed bug free. Nobody has published one. Virginia Cooperative Extension suggests a follow up inspection within a week confirming that aggregation sites are empty, no live bugs have been seen and nobody has been bitten.

The problem is that visual inspection is least reliable in exactly the condition that follows a treatment, which is a small number of surviving insects. Research found visual inspection detected only around half of infestations in lightly infested apartments. In another study, visual inspection found an average of 12 bed bugs per apartment, and a trap placed in the same units afterward collected 15 more.

Field studies assess outcomes at eight to twelve weeks, with monitoring often extending to 22 or 28 weeks. A treatment declared successful at one or two weeks has no support in the research timeline, because eggs laid before the treatment can still hatch and because the surviving population is below the level visual inspection reliably finds.

This is the part of the process we are involved in, so treat the next sentence with appropriate suspicion. A canine sweep two to four weeks after treatment is a different kind of check from a visual walkthrough, because the dog is working from scent at harbourage points rather than from what is visible on a surface. What makes it useful is not the dog, it is that the person running it is not the person who did the treatment and does not get paid more if the answer is good or bad.