How safe are chemical tick treatments for dogs?

For years, whenever possible, I have preferred natural ways of protecting my dogs against ticks, and Oscar is the main reason behind that choice. When he was a puppy, I used Bravecto tablets because I had seen with Faddy how effectively they kept ticks away. Oscar received the tablet twice. The first time I did not notice anything unusual. The second time he lost his appetite and became itchy all over. I cannot prove in retrospect that Bravecto was definitely the cause, but the timing was strong enough for me that I did not want to try the same or a similar product on him again.

Since then I have relied more on garlic granules, manual tick checks and natural repellents before going into areas with a higher tick risk. My experience has been good. Experience is not the same as clinical evidence, but the absence of evidence does not make a real experience meaningless either. The point of this article is not to say that every chemical tick product harms every dog. The point is to look honestly at the risks that are often glossed over inside the convenient phrase “safe and effective”.

Is “safe” even the right word?

The FDA says that it considers isoxazoline products safe and effective for the majority of dogs and cats, while the same agency also requires a warning about possible neurological adverse reactions. For me, it is important to see both statements at the same time rather than take the first and forget the second.

A marketing authorisation essentially means that the regulator has judged the benefit-risk balance of a medicine to be acceptable for a particular indication and according to the authorised directions for use. It does not mean that the product is risk-free, that everything is known about long-term repeated use, or that every individual dog will react in the same way. Personally, I do not simply call such products “safe”, especially when a drug class carries an official neurological warning and a very large number of adverse-event and other post-marketing reports has accumulated.

These substances act through the nervous system

Isoxazolines such as fluralaner (Bravecto), afoxolaner (NexGard) and sarolaner (Simparica) are insecticides and acaricides, meaning substances that kill fleas, ticks and other arthropods. Their target is the parasite’s nervous system. Fluralaner, for example, blocks GABA- and glutamate-gated chloride channels; with afoxolaner, the result is described as uncontrolled hyperexcitation of the parasite’s central nervous system followed by death.

So the word “neurotoxin” has not appeared out of nowhere: their antiparasitic action is neurotoxic to the target organism. At the same time, precision matters. Isoxazolines are designed to act more strongly on arthropod receptors than on mammalian receptors. That selectivity is why they can be used in dogs at all. But the FDA’s official warning about muscle tremors, ataxia and seizures shows that the mammalian nervous system is not completely outside the picture when it comes to possible adverse effects.

Adverse-event reports: not proof of causation, but not a meaningless signal

One distinction is essential here. An adverse-event report means that an event occurred after a medicine was used and that someone reported it. It does not automatically prove that the medicine caused the event. Nor can these reports be used to calculate the true incidence of adverse effects, because there is no precise denominator, reports may be incomplete, and very widely used products will inevitably generate more reports.

But the opposite extreme would be to say that because each individual report does not prove causation, a large volume of reports means nothing. It does mean something: pharmacovigilance exists specifically to detect post-marketing patterns that small pre-authorisation studies may not reveal.

A 2020 analysis of isoxazoline safety brought together the FDA and EMA data available at the time and already showed that post-marketing reports were far from a handful. Today, the public EudraVigilance Veterinary database contains tens of thousands of suspected adverse-event cases associated with several commonly used tick-control active ingredients. This does not prove that the medicine caused every case, but it is exactly the kind of real-world safety signal that pharmacovigilance is designed to monitor.

On 23 September 2026, I reviewed the main active ingredients currently or historically used for tick control in dogs through the public EudraVigilance Veterinary substance search. Below I deliberately use only the “Number of cases” measure. “Animals treated” and “Animals affected” are not the denominator of all animals treated on the market, so they cannot be used to calculate an adverse-event percentage.

EudraVigilance Veterinary: active-ingredient reports as of 23 September 2026

These are substance-level veterinary reports of suspected adverse events across all species in the database unless the view has been separately filtered by species. I therefore do not present them as “proven adverse effects in dogs” or as a ranking of risk. They show reporting volume, not causation or incidence.

Active ingredient	Number of cases	Class / note
Afoxolaner	94,071	isoxazoline
Amitraz	889	formamidine
Deltamethrin	8,155	pyrethroid
Dimpylate	283	organophosphate
Fipronil	73,953	phenylpyrazole
Flumethrin	81,570	pyrethroid
Fluralaner	81,255	isoxazoline
Imidacloprid	106,375	neonicotinoid
Lotilaner	13,895	isoxazoline
Permethrin	14,713	pyrethroid
Permethrin (25:75)	358	pyrethroid; separate EudraVigilance entry
Permethrin (40:60)	8,046	pyrethroid; separate EudraVigilance entry
Phoxim	272	organophosphate
Propoxur	391	carbamate
Pyriprole	461	phenylpyrazole
Sarolaner	97,441	isoxazoline
Tetramethrin	157	pyrethroid

 

Important: these figures must not simply be added together. The same case may involve a combination product and therefore appear under more than one active ingredient. A higher report count also does not automatically mean a higher individual risk: usage volume, time on the market, reporting culture and product type all affect the number of reports.

At the same time, it is difficult to dismiss these safety signals as trivial. Afoxolaner, sarolaner, fluralaner, fipronil, flumethrin and imidacloprid each have tens of thousands of suspected adverse-event cases in their public substance views. Neurological disorders are also visible as a reaction category for many ingredients. This is not a risk percentage, but it shows that neurological reactions are not merely a theoretical sentence in a package leaflet.

Why did the number of reports jump in recent years?

EudraVigilance itself warns that data collected before and after January 2022 should not be compared one-to-one. Before 28 January 2022, mandatory reporting focused mainly on serious adverse events. When the new EU veterinary medicines legislation took effect, reporting became mandatory for both serious and non-serious suspected adverse events. According to the EMA’s 2022 annual report, a backlog of earlier reports was also entered into the system that year, contributing to a large increase in report numbers.

So it is accurate to say that reporting of adverse events has risen sharply in recent years. I would not, however, say that the rising graph automatically proves the medicines themselves have become more dangerous. The increase is influenced at the same time by broader reporting requirements, greater awareness, product usage and potentially genuine safety signals. The Bravecto annual graph shows a very clear jump from 2022 onward, but the 2022 reporting change must always be mentioned when interpreting it.

And not every adverse event is reported

One of the biggest weaknesses of spontaneous pharmacovigilance is under-reporting. A 2024 study based on real clinical records from UK small-animal practices identified 827 suspected adverse drug reactions, and in roughly 90% of those records there was no indication that the case had been reported to the regulator or the marketing authorisation holder. An earlier European survey of veterinarians found a similar order of magnitude: veterinarians estimated that they saw about ten times more adverse reactions than they formally reported.

This is why the statement that “perhaps only around one tenth of suspected adverse reactions reach the official system” is defensible in veterinary medicine as an approximate order of magnitude, not as a universal constant. I would not automatically multiply every EudraVigilance number by ten because studies, countries and reporting systems differ. But it does mean that public report counts probably describe only part of the suspected reactions actually observed.

The official neurological warning for isoxazolines

Since 2018, the FDA has warned that isoxazolines have been associated with neurological reactions including muscle tremors, ataxia or loss of coordination, and seizures. Seizures have also been reported in animals without a previous seizure disorder. This warning remains current in 2026 and the same class warning has been added to product labelling.

In addition to the officially listed neurological reactions, post-marketing reports describe vomiting, diarrhoea, reduced appetite, lethargy, itching and other signs. Dr Judy Morgan also describes behavioural changes, disorientation, an unsteady gait, sensitivity to touch, and urinary or faecal incontinence among reactions she has seen in clinical practice. These are her clinical observations, not incidence rates from controlled studies, but they are still symptoms an owner can reasonably watch for after treatment.

If a dog develops a seizure, tremors, ataxia, marked lethargy, repeated vomiting, unusual behaviour or another clear change after a product is used, do not simply wait for the next dose. Contact a veterinarian, and the suspected adverse event is also worth reporting formally. Keep in mind that signs may sometimes appear later rather than immediately after administration, including during the product’s period of activity.

A practical difference with oral products: once swallowed, they cannot be washed off

For me, one practical risk with oral products is that after the active ingredient has been absorbed, it cannot be removed from the dog’s coat or skin. Dr Judy Morgan makes the same distinction: if there is a strong reaction to a spot-on product, at least some of the product can be washed off quickly; with an oral medicine, that option does not exist.

The advantage of a long-acting medicine is convenience and consistent parasite protection. The same property can be a disadvantage if an adverse reaction occurs. With oral Bravecto, for example, fluralaner concentrations remain measurable for a long time after administration. This does not mean that an adverse effect necessarily lasts for that entire period, but it does mean that systemic exposure does not end the next morning.

Tick collars and Seresto: case reports cannot simply be brushed aside

A very large number of adverse-event and death reports accumulated in the United States in relation to Seresto collars. The EPA analysed around 1,400 death reports submitted between 2016 and 2020. For most deaths, the agency did not find enough evidence to identify the product as a definite cause, and many reports lacked sufficient information. That nuance needs to be stated honestly.

But the story did not end with the sentence “causation has not been proven”. The EPA’s own scientific review found moderate and severe non-fatal cases in which symptoms improved after the collar was removed and returned when the collar was used again. The EPA therefore required additional warnings and better adverse-event reporting. More importantly, the EPA Office of Inspector General concluded in 2024 that the agency’s existing approach did not provide sufficient assurance that the collars did not pose an unreasonable risk to pet health. For me, that remains an important safety signal even when causation cannot be proven in every individual case.

Does the active ingredient in a tick collar stay only on the coat?

It is not accurate to say that Seresto works simply by first sending the chemical into the dog’s bloodstream. Its main mechanism is the continuous release and distribution of the active ingredients over the skin and coat. But it would be equally inaccurate to say that the ingredients never enter the dog’s body at all. In official pharmacokinetic data, imidacloprid reached the systemic circulation at least transiently, while flumethrin was generally not measurable in blood.

For me, the reasonable wording is this: the collar’s action does not primarily depend on a high systemic blood concentration, but a substance placed on or distributed through the skin is not automatically completely separated from the body. Absorption depends on the particular active ingredient and formulation.

Spot-on products: the dog, the human and the environment all receive exposure

One warning from the older version of this article has gained much stronger evidence. In a 2024 study, fipronil and imidacloprid were detected in 100% of wash-off samples from dog bathing, bedding washing and owners’ handwashing. Owner handwashing continued to carry some of the active ingredient into wastewater for at least 28 days after application. A 2026 study of Welsh rivers found imidacloprid and fipronil much more frequently in urban samples, with predicted no-effect concentrations exceeded in a substantial proportion of urban samples.

The product labels themselves also show that human exposure is not treated as nonexistent. The US label for Bravecto topical for dogs, for example, says the application site should not be touched, or touched by children, for two hours. NexGard Combo topical for cats specifies four hours or until the site is visibly dry. Several European labels for fipronil spot-ons say that treated animals should not be handled until the application site is dry and advise against allowing freshly treated animals to sleep in the owner’s bed, particularly a child’s bed.

That does not by itself prove that the product is unacceptably dangerous for a dog. But to me it does not sound like a substance that should be treated as harmless cosmetics either. It is a biologically active pesticide and deserves to be handled with the corresponding respect.

Fipronil: the old cancer warning needs context, not deletion

In a historical US EPA risk assessment, fipronil was classified as Group C, or a “possible human carcinogen”. The classification was based on an increased incidence of thyroid tumours in rats. This does not mean that a canine spot-on product is a proven cause of cancer in people or dogs, and later EPA cancer-risk frameworks use different descriptors. But the older statement does not need to be erased. The accurate approach is to explain what the classification was based on and what it does not allow us to conclude.

If you choose chemical tick control, what should you do about the dose?

The conventional dosing advice is simple: use the specific product according to the authorised weight range in its leaflet. It is not a good idea to arbitrarily reduce the dose yourself because too little may provide inadequate protection, and not every tablet is designed to be split.

At the same time, integrative veterinary medicine raises a legitimate question about very broad weight bands, where a dog at the lower end of a range can receive considerably more mg/kg than a dog at the upper end. Dr Judy Morgan has specifically drawn attention to this issue in smaller animals in her interviews.

Interestingly, a 2026 study evaluated a shorter-acting formulation with a lower fluralaner dose: 10-22.5 mg/kg maintained very high efficacy against fleas and ticks through 49 days in a controlled study. This is a different formulation with a different registered dosing schedule, so it does not prove that an existing Bravecto tablet can simply be cut into a smaller dose at home. What it does show is that “more active ingredient and longer duration” is not the only possible design, and the question of lower systemic exposure is entirely legitimate.

If my own dog were right at the bottom of a weight band, I would ask the veterinarian specifically which registered option gives the lowest necessary exposure, whether a shorter-acting alternative exists, and whether that particular tablet or formulation allows any dose adjustment at all. I would not automatically choose the longest-acting option purely for convenience.

Can the body be supported?

In integrative veterinary medicine, liver and antioxidant support is often used when there is concern about chemical burden or a suspected adverse reaction. Dr Judy Morgan has written about milk thistle, NAC, glutathione, chlorella and other supportive options. These recommendations belong to integrative clinical practice, and we do not have good clinical evidence that they “neutralise” the neurological risk of an isoxazoline.

If a dog develops a genuine adverse reaction, the first step is veterinary assessment. Integrative support can be used alongside that, not instead of urgent care.

Natural tick control: my preference does not mean “put any oil on your dog”

I still prefer a more natural strategy for my own dogs whenever possible. It does mean more work: daily body checks during tick season, fast tick removal, making deliberate choices in higher-risk areas, using garlic granules when appropriate, and applying a topical repellent before going into woods or tall grass.

I have written separately and in detail about garlic. I do not treat it as guaranteed tick protection, but my own practical experience has been good, and integrative veterinary medicine uses garlic as one layer of parasite control. Under very high tick pressure, I would not rely on a single natural method alone.

My DIY tick spray for dogs

This is one DIY recipe I use:

• 240 ml clean water
• 120 ml organic unfiltered apple cider vinegar
• 20 drops neem oil
• if needed, 5 drops of one essential oil: lemon, lemongrass, eucalyptus or geranium

Shake well before use. Spray onto the dog’s coat before entering a higher-risk tick area, especially the legs, underside and neck. I do not spray the face directly. If needed, I spray the mixture onto my hands and gently smooth it over the fur, avoiding the eyes, nose and mouth.

Dilution matters greatly with essential oils, and “natural” does not automatically mean every dog will tolerate every oil. I always start on a small area and watch the skin, breathing and behaviour. For a very young, sick, neurologically affected, pregnant or particularly sensitive dog, it is sensible to discuss the mixture first with a veterinarian or a practitioner familiar with veterinary phytotherapy. Cats require much greater caution because their metabolism differs from that of dogs and several essential oils may be unsuitable for them. A dog recipe should never automatically be transferred to a cat, because both dilution rules and suitable oils may differ.

A very similar recipe has also been published under Dr Karen Becker’s name using water, apple cider vinegar, neem and catnip oil, plus a small amount of one essential oil. The key point is that a spray like this is a repellent layer, not a guarantee. I still check my dogs after a walk.

Do ticks “like” a healthier dog less?

This idea is a genuine part of the integrative perspective. Based on his clinical experience, Dr Peter Dobias has written that dogs in good overall health appear to be less susceptible to fleas and other parasites. At the same time, he is very clear that even a completely healthy dog can pick up fleas or ticks under heavy parasite pressure, and good nutrition alone is not parasite control.

Ticks locate a host through CO2, heat, odours, moisture and vibration. We do not have a good controlled canine study proving that “a healthier immune system means a tick will not attach”. My way of putting it is this: according to the clinical experience of integrative veterinarians, an animal’s overall health may influence parasite burden and susceptibility, but it should not be used as the only tick-control method. For me, diet, the skin barrier, the microbiome, stress and overall health are still part of the whole picture, just not an invisible tick shield.

Ticks themselves are also a real risk

None of the above means that the danger from ticks should be minimised. In Estonia and elsewhere in Europe, ticks can transmit Lyme borreliosis, anaplasmosis, babesiosis and other diseases. Every dog therefore requires consideration of two risks at once: the risk from parasites and tick-borne disease, and the risk associated with the control method being used.

A dog that walks through woodland and tall grass every day is not in the same situation as a dog that mainly stays in maintained urban areas. The choice also changes if the dog has previously reacted to a tick product, has a seizure disorder, or has an owner who can perform a very thorough tick check after every walk.

Conclusion

Chemical tick treatment is not automatically my first choice, and I do not simply call these products “safe”. They are biologically active insecticides and acaricides, some of which act systemically and some of which spread over the skin and coat. With isoxazolines, a risk of neurological adverse effects is officially recognised; with spot-ons and collars there are also questions of exposure for animals, people and the environment.

At the same time, I do not make the opposite mistake and claim that every adverse-event report proves medicine-induced harm or that every dog will experience side effects. My concern is when risks are minimised with the word “safe” as though warnings, pharmacovigilance, product information and individual reactions did not exist.

For my own dogs, Oscar’s experience tips the balance towards natural repellents, daily tick checks and deliberate risk management. For another dog, medication may be justified in some situations, particularly when tick pressure or tick-borne disease risk is very high. To me, the decision should be as individual as possible: how much protection does this dog really need, what is the dog’s health background, what is the history of previous reactions, and how little chemical exposure can achieve the necessary goal?

An informed choice means that the owner knows about possible and potentially serious adverse effects before using the medication, not only after their pet experiences one.

Checked sources:

• FDA – Fact Sheet for Pet Owners and Veterinarians about Potential Adverse Events Associated with Isoxazoline Flea and Tick Products
• FDA – Animal Drug Safety Communication: neurologic adverse events and isoxazolines
• FDA / DailyMed – Bravecto chewable, mode of action and pharmacokinetics
• FDA / DailyMed – Bravecto topical, human contact warning and mode of action
• FDA / DailyMed – NexGard, mode of action
• Palmieri et al. – Survey of canine use and safety of isoxazoline parasiticides, Veterinary Medicine and Science, 2020
• US EPA – Seresto Pet Collar Review
• US EPA Office of Inspector General – Seresto report, 2024
• Seresto pharmacokinetics – imidacloprid transiently reaches systemic circulation
• Perkins et al. – Down-the-drain pathways for fipronil and imidacloprid applied to dogs, 2024
• Occurrence and sources of veterinary ectoparasiticides in Welsh rivers, 2026
• US EPA – Fipronil risk assessment / Group C possible human carcinogen classification
• Scott & Campos – short-acting lower-dose fluralaner formulation, 2026
• Dr Judy Morgan – Side Effects from Common Flea and Tick Products
• Dr Judy Morgan – Six Ways to Improve Your Pet’s Health Naturally
• Dr Peter Dobias – Can feeding a raw diet prevent fleas?
• Dr Peter Dobias – Natural Tick Prevention and Lyme Disease in Dogs
• Dr Jean Hofve – Dr Becker’s Homemade Flea & Tick Repellent Spray
• EudraVigilance Veterinary – Data source
• EudraVigilance Veterinary – Search by substance
• EMA – 2022 Annual Report: veterinary adverse-event reporting and the 2022 reporting change
• Recording of suspected adverse drug reaction reporting in veterinary free-text clinical narratives, 2024 (PMID 38441325)
• Veterinary pharmacovigilance in Europe: a survey of veterinary practitioners, 2017 (PMID 28848652)

Note: an adverse-event report is a pharmacovigilance signal, not automatic proof of causation or incidence. Substance-level EudraVigilance figures must not be added together into a risk percentage or used as a safety ranking. At the same time, strong under-reporting is documented in veterinary pharmacovigilance, so the public report count probably does not capture every suspected reaction that is actually observed. The experience and recommendations of integrative veterinarians are clearly separated here from regulatory data and controlled studies.

---

---