When to Treat for Varroa: The Timing That Decides Winter

Two curves run through every beekeeping season, and varroa management is the art of keeping them apart. The bee population curve rises from spring, peaks at the flow, and contracts toward the compact winter cluster. The mite curve rises with it but keeps rising after the bee curve turns down, because mites reproduce in brood and late-season colonies still have brood while their adult populations shrink. The result is the deadly scissors of September: mites per bee can quadruple in the exact weeks the colony is rearing the winter bees that must live until March. A colony can look boiling-strong at the honey harvest and be doomed already, and the beekeeper who treats when the colony finally looks sick, in October, is treating a month after the verdict came in.

This is why the same product succeeds and fails in different hands. The strips do not care whether you believe in them; they kill mites whenever applied. But killing mites in October cannot repair winter bees that were parasitized as pupae in September, cannot restore the fat bodies and virus resistance they lost, and so the colony dwindles and dies in January with clean mite counts, honey in the comb, and a bewildered keeper. The Honey Bee Health Coalition's Tools for Varroa Management, the closest thing US beekeeping has to a consensus playbook, is organized around exactly this point: decisions run on monitoring and the seasonal phase of the colony, not on the shelf. Product choice matters, and our treatment roundup ranks the field, but read this page first, because a mediocre product on time beats the perfect product late, every single year.

Everything on this page keys off one number, mites per 100 adult bees, and the only way to get it reliably is a wash: shake about 300 bees (a level half cup) from a brood-nest frame into alcohol or soapy water, agitate, and count the dislodged mites; divide by three and you have your percentage. Confirm the queen is not in your sample before it goes in the jar. The purpose-built jar makes this a five-minute chore: the Mann Lake Varroa EasyCheck ($30.50) is the standard, a two-chamber wash cup with a marked fill line, and the technique plus its gentler sugar-roll cousin are stepped through in the mite monitoring tools guide. Sample from a brood frame, because that is where mites concentrate; a sample from the honey super flatters the hive and lies to you.

A word on the popular alternatives. Sticky boards under a screened bottom count natural mite fall per day, and they are fine as a trend line and as a treatment-verification tool, but their conversion to actual infestation is loose enough that extension guidance, including the University of Minnesota's monitoring protocol, treats washes as the decision-grade method. The sugar roll returns the bees alive at a modest cost in accuracy, a fair trade for small-scale keepers who cannot bring themselves to sacrifice the sample; do it consistently and it steers fine. Eyeballing, watching for mites on adult bees, is worthless in the cruel technical sense: mites ride between the abdominal plates and inside cells, and by the time you can see mites on bees, or bees with deformed wings, the infestation is severe and the virus damage under way. Count monthly, the same way, every time, and write it down; the trend across months is worth more than any single number.

The consensus action threshold from the Honey Bee Health Coalition and extension programs is 2 to 3 mites per 100 bees, but the working number tightens and loosens with the season, because the same count means different things in May and August. Here is the threshold table we run, on a US temperate baseline:

Month	Act at (mites/100 bees)	Why
March / April	3	Populations exploding; a spring count above 3 means a bad August is already loading. Treat before supers.
May / June	2 to 3	Brood season doubling underway. Supers complicate product choice; see the summer section.
July	2	The predictor count. At or near 2, an August treatment is certain: buy product now, plan the window.
August	2, or treat on principle near it	Winter bees rear in September. This is the count that decides winter.
September / October	2	Reinvasion season: robbing imports mites from collapsing hives. Verify August held; retreat on rebounds.
Nov / December	Any meaningful count	Broodless oxalic window: cheap, easy, near-total exposure. Fire it regardless in most operations.
Jan / February	Monitor only	Too cold to wash in the North; sticky boards if screened bottoms allow. Southern yards: keep counting.

Two habits make the table work. First, thresholds are triggers, not targets: the goal is staying under them, and a colony that keeps riding the line is telling you about its neighbors, its genetics, or your intervals. Second, every count goes in the notebook with a date; the treatment decisions in the rest of this page are made by comparing this month's number to last month's, and memory is not a monitoring method.

If this page had to be one paragraph, it would be this one. Sometime in late August through September, the colony begins rearing winter bees: a physiologically distinct generation with enlarged fat bodies, built to live months instead of weeks and to carry the colony, its warmth, and its spring restart through to March. They are reared once. Varroa feed on the fat bodies of developing pupae, the exact tissue winter bees need, while vectoring deformed wing virus into them, and pupae parasitized in September emerge as winter bees with a fraction of their lifespan. The colony then dwindles through December and dies around January, and the autopsy shows the cruel signature: a tiny cluster, plenty of honey, few dead mites, because the treatment the keeper finally applied in October worked perfectly on mites and did nothing for bees already damaged in the cradle. University extension programs converge on the same instruction, treat before winter-bee rearing, not after; Penn State's varroa management guidance and the Coalition's tools both put the emphasis on late summer for exactly this reason.

So the operational rule: count in early August, and treat at or approaching 2 per 100, with the treatment finished before September's rearing is underway. In the northern tier, where winter bees start earlier, shift the whole operation two to three weeks earlier, into late July; in the South, later, but do not relax, because your reinvasion season is longer too. Which product? August is usually hot, so the choice is a temperature negotiation, next section, but the classic August call where highs cooperate is a formic product like Formic Pro ($30.00, two doses), for one decisive reason: formic acid penetrates brood cappings, per its label, and kills mites where they reproduce, which no other mainstream chemistry does while honey supers are a consideration. The full August decision tree, formic versus thymol versus amitraz by scenario, is in the treatment roundup; what is not negotiable is the month itself. August treatment is winterizing, the first and heaviest item on the list in the winterizing guide.

Every varroa chemistry is temperature-sensitive somewhere: too cold and vapors do not vaporize, too hot and organic acids turn from treatment to harm. The label window on the box is a legal boundary and a biological one, and the single most common treatment failure after bad timing is running a product outside it. Here are the mainstream windows, hedged exactly as the labels hedge them; read your box, because formulations and labels change:

Product (class)	Label temperature window	Reaches capped brood?	Supers on?	Typical duration
Formic Pro (formic acid)	About 50 to 85F, per label	Yes, per label	Yes, per label	14 to 20 days
Apiguard (thymol)	About 60 to 105F, per label	No	No; remove supers, per label	2 x 2 weeks
Apivar (amitraz)	Broad; brood-season use, per label	No; relies on 42-day exposure	No; remove supers, per label	42 to 56 days
Oxalic acid vapor/dribble	Applied on mild days; works in cool weather	No	Per current label	Minutes; repeat per label
VarroxSan (oxalic, extended-release strips)	Broad brood-season window, per label	No; extended exposure spans brood cycles	Per label	Up to about 56 days

The practical reading of the table: formic is the precision August weapon where a 14-to-20-day stretch of highs under about 85F exists, with known trade-offs (brood pause, occasional queen loss, strong fumes) that the label mitigations address. Thymol is the hot-yard alternative, tolerant up over 100F, at the cost of pulling supers and a month of treatment. Amitraz is temperature-flexible and gentle on bees but needs its long window plus super removal, and it is the class where resistance reporting is loudest, see rotation below. Oxalic in its flash forms is the cool-weather and broodless specialist, and in its extended-release strip form becomes a brood-season player by brute persistence. Match the forecast to the label before you buy, not after; an August heat dome has stranded many a box of formic strips in the shed while the mites compounded.

Spring varroa work is about protecting two things: the buildup and the honey. A colony that wintered with mites starts the doubling clock from a higher floor, and a spring count above threshold means intervening before supers go on, while every product is still legal to use and the flow has not started the clock. The spring-legal menu is the full menu, so choose by forecast: formic in the mild window, thymol as it warms, amitraz if you accept its 42-day runway ending before supering, and for the earliest intervention of all, a broodless or near-broodless oxalic shot in late winter, when the cluster is small and cappings are few, an underrated window we flag in the calendar below.

Once supers are on, the rules change and the menu shrinks: most products, thymol and amitraz included, are off the table with marketable honey stacked, per their labels, and formic stands nearly alone in being labeled for use with supers on. This is the strategic reason the spring count matters so much: it is your last cheap decision before the flow locks the toolbox. A colony crossing threshold in June with supers on faces the choice between a formic treatment in flow heat, pulling supers early, or waiting and gambling the August window, and none of those is as good as the April treatment that would have prevented the whole dilemma. Spring is also when drone-brood culling earns its keep as a non-chemical assist for the IPM-minded, mites prefer drone brood, and freezing a sacrificial drone frame monthly skims the mite population, slowing the doubling without touching the honey; the technique and its hardware live in the monitoring guide's IPM section.

Twice a year, varroa's fortress empties. Every mite in a colony is either reproducing under a brood capping, untouchable by most chemistries, or riding exposed on adult bees. Whenever brood approaches zero, the fortress stands empty and every mite is in the open, and a single well-timed oxalic acid application in a broodless colony achieves the kind of kill percentages no brood-season treatment approaches. The premier broodless window is late fall: after a hard cold stretch, typically late November into December in the baseline region, brood shrinks to nothing or nearly so, and a single oxalic vaporization or dribble on a mild day, cluster loose enough for vapor to reach through, is the cheapest, fastest, most effective treatment of the whole year. It cleans up whatever August missed and whatever fall robbing imported, and it starts the spring buildup from a near-zero mite floor. Most experienced cold-country keepers fire this shot annually on principle, count or no count.

The second broodless window is one you create or catch: a swarm's or package's first week. A newly hived swarm or installed package has zero capped brood until day eight or nine, and a single oxalic treatment in that window knocks the imported mite load down before the first cappings seal mites back into the fortress; the same logic applies to a colony that has been requeened with a long brood break, or a walk-away split during its queenless gap. Swarm aftercare in the swarm guide leans on exactly this. Application mechanics matter in the cold window, vapor reaches a winter cluster better than dribble in most configurations, dosing and repetition follow the current oxalic label, and the equipment tiers from 12-volt wands to cordless pro units are ranked in the vaporizer guide. Respiratory protection is non-negotiable with vapor: an acid-rated respirator, not a dust mask, per every label and every extension safety sheet ever written.

The nastiest surprise in varroa management is the clean colony that spikes in October, and it has a name: reinvasion, or mite immigration. As untreated colonies in flying range collapse through fall, their mite-loaded bees drift and beg into healthy hives, and your own foragers rob out the dying colonies and carry mites home like groceries. Studies and painful experience agree a colony can import a whole season's worth of mites in a few robbing weeks, which means an August treatment, verified and successful, does not close the books; September and October counts do. Keep washing monthly through fall, and treat rebounds promptly with whatever the temperature still allows, formic early in fall, then oxalic forms as the weather cools and broodlessness approaches.

This is also the section where varroa management stops being a private matter. A collapsing hive is a mite bomb that shells every apiary in two miles, which is why the ethic among serious keepers is blunt: treat, monitor, or requeen toward resistant stock, but do not keep untreated, unmonitored colonies stumbling toward autumn collapse in a neighborhood with other beekeepers. It is also the argument for reduced entrances and robbing screens through fall, hardware covered in the robbing screen guide: less robbing traffic is fewer imported mites, on top of everything else it prevents. And it explains a pattern that otherwise looks like bad luck: the keeper whose colony dies of varroa despite treating on time, because a feral colony in a wall two houses over collapsed in October. You cannot control the neighborhood; you can count in October and catch what it sent you.

Varroa's generation time is measured in weeks, which makes it an evolution engine, and the history of varroa control is a graveyard of products that worked until they did not: the older synthetic classes bred widespread resistance years ago, and resistance reporting around amitraz, the workhorse synthetic, has grown loud enough that no one should treat it as permanent. The countermeasure is the same as in every pest system: rotate chemistry classes, never running the same class back-to-back season after season. The organic acids and thymol carry a structural advantage here, their modes of action are considered hard for mites to evolve around, which is why a rotation built on formic, thymol, and oxalic, with a synthetic held in reserve, is the pattern most extension IPM guidance sketches and the one we run.

A workable three-year rotation for a temperate yard: year one, formic in August with an oxalic broodless shot in December; year two, thymol in August (or extended-release oxalic strips like VarroxSan spanning late summer), oxalic in December; year three, if counts or reinvasion demand a heavier hand, the synthetic, then back to the acids. Alongside rotation, the practice that actually detects resistance: recount one to two weeks after every treatment ends. An August formic treatment should crater the wash count; a post-treatment count that barely moved means a misapplied product, an out-of-window application, reinvasion, or resistance, and every one of those diagnoses demands a second treatment with a different class, now, not a shrug. The keepers who get burned are not the ones who chose the wrong product; they are the ones who never checked whether the product worked.

The whole system on one table, US temperate baseline (zones 5 to 7). Northern tier: shift the August entries two to three weeks earlier and expect a longer, deader broodless window. Gulf South: see the next section, because your table has no off month.

Month	Monitoring	Treatment action	Notes
January	Sticky board only, if any	None (or late oxalic if missed in Dec and a mild day allows)	Do not open the hive; see the winter rules.
February	Sticky board	Near-broodless oxalic possible late month in mild regions	Brood restarting; window closing.
March	First wash when 55F+ allows	Treat above 3/100, any legal product by forecast	Pre-super window opens.
April	Monthly wash	Treat at threshold, finish before supers	Drone-comb culling starts.
May	Monthly wash	Formic only if supers on and treatment forced	Flow begins; toolbox narrows.
June	Monthly wash	Hold under threshold; formic if forced	Doubling season in full swing.
July	The predictor wash	Order August product now if count near 2/100	Plan the temperature window.
August	Early-month wash	The knockdown: formic in window, thymol if hot	Finish before September rearing. The year's hinge.
September	Post-treatment verification wash	Retreat with a different class if count did not crater	Winter bees rearing now.
October	Reinvasion wash	Treat rebounds; cool-tolerant options	Robbing season imports mites.
November	Board or final mild-day wash	Broodless oxalic when cold has emptied the brood nest	The cleanup shot.
December	None	Broodless oxalic if not yet fired	Vapor on a mild day; gear in the vaporizer guide.

Print it, argue with it at your club, adjust it to your latitude, and then actually follow it; a mediocre calendar executed every month outperforms a perfect one consulted in October.

Gulf South, Florida, southern California, south Texas (zones 8 to 10): your colonies brood essentially year-round, which deletes the two pillars northern keepers lean on: there is no deep broodless window for a near-perfect oxalic shot, and no winter shutdown of mite reproduction. The doubling clock runs all twelve months, reinvasion season is longer, and the operational translation is simple and unforgiving: monitor monthly, all year, and be ready to treat in any month the count demands, choosing by temperature (thymol and extended-release oxalic strips carry more of the load in your heat, formic in your milder shoulders). Your best approximations of a broodless window are manufactured ones, requeening breaks and splits, and the brood-spanning persistence of extended-release oxalic is particularly suited to your problem. Treat the August-window principle as a September-October window: your winter bees are later, but they are still real, and so is their September vulnerability.

Northern tier and high elevation (zones 3 to 4): compress everything: winter bees start rearing in August, so the knockdown belongs in late July, the fall verification in August, and your compensation is the best broodless window in the country, long, deep, and reliable, where a single November oxalic treatment approaches total knockdown. Arid West: watch the temperature ceiling, formic's 85F limit is your binding constraint most of the season, making thymol and oxalic forms your workhorses, and monsoon-season humidity swings affect formic release rates, another reason the label's application conditions matter. Wherever you keep bees, the local club knows the local mite calendar, when the collapses start, when robbing peaks, and their dates outrank this page's baseline, exactly as we said about every calendar in the first-year guide.

Every fall, colonies die whose keepers treated for varroa, and the failure modes repeat so reliably they are worth naming as a checklist. Treating late: the October treatment after the September damage; the whole first half of this page. Treating blind: no counts before (was treatment needed? was the timing right?) or after (did it work?), which converts treatment from a control system into a ritual. Ignoring the label window: formic in a 95F week harms bees and queens; thymol at 55F does approximately nothing; both read as the product failing when the calendar did. One-and-done thinking: the verified August treatment followed by zero fall counts, straight into an October reinvasion spike nobody measured. Same class forever: the resistance section, ignored until the year the strips stop working. And the quiet one, skipping the December oxalic because the colony seems fine: the broodless shot is the cheapest insurance in beekeeping, and spring starts from whatever floor you set in December.

The system that avoids all six fits on an index card: wash monthly, treat at 2 to 3 per 100 by the label window, own August, verify every treatment, guard the fall, fire the broodless shot, rotate classes, write it all down. The products themselves are solved problems, ranked and priced in the treatment guide, with the counting hardware in the monitoring guide and the winter equipment in the vaporizer guide. What no guide can sell you is the calendar discipline, and that is the actual treatment. Varroa is not beaten by chemistry. It is beaten by beekeepers who know what month it is.