Ronidazole is the only treatment shown to eliminate this infection, it is not licensed for cats, and the dose recommended today is not the dose most of the internet still prints. This page lays out the evidence in order, including what has already been tried and failed.
Before any drug: in a study following cats with this infection, diarrhea usually resolved on its own within about two years, and cats that were treated took longer to resolve.1 That is not an argument against treating — it is an argument against treating reflexively.
The distinction that matters most: the diarrhea stopping is not the infection ending. Untreated cats generally remain carriers, sometimes for life, and the diarrhea can return after stress. So the trade-off is not "sick versus well" but "risk of drug toxicity versus risk of a persistent carrier in your household."
The owner's path walks through that decision →
If you have read that the dose is 30–50 mg/kg twice daily, you have read something that was correct in 2006 and was withdrawn in 2014. The timeline is the only way to recognize stale advice, so here it is.
| Year | What was recommended | Why it changed |
|---|---|---|
| 2006 | 30–50 mg/kg twice daily for 14 days | Cleared infection in 10 of 10 experimentally infected cats, PCR-negative 21–30 weeks later. A 10 mg/kg dose failed in all cats.2 |
| 2011 | — | Pharmacokinetic study: oral absorption ≈99.6%, terminal half-life ≈10.5 h, drug still detectable 48 h after one dose. Twice-daily dosing predicted to accumulate.3 |
| 2014 | 30 mg/kg once daily for 14 days | The original authors withdrew the twice-daily recommendation in a published letter, citing accumulation and risk of neurological toxicity.4 |
| 2024 | — | A repurposed gold-salt drug, auranofin, killed the organism in vitro and was ineffective in naturally infected cats. Published as a dead end.5 |
Ronidazole is a nitroimidazole, related to metronidazole. It is not approved by the FDA for use in cats — or in any companion animal — and it has been banned in food animals over concerns about human health effects. It is prescribed off-label and compounded.
Its narrow safety margin is the central problem. Reported neurological signs include lethargy, ataxia (a wobbly, drunk-looking walk), tremors of the limbs, facial tremor, agitation, hyperesthesia, abnormal eye movement, and seizures. Signs typically begin at least three days into treatment and usually resolve one to four weeks after the drug is stopped.
Ronidazole is also considered contraindicated in pregnant or lactating queens, and kittens under twelve weeks are usually given a reduced dose because of immature metabolism.
Ronidazole is a suspected human reproductive and carcinogenic hazard; that is the reason for the food-animal ban, not a precaution invented for cats. Compounded powders are the highest-exposure form. Shelters, rescues and fosterers should treat it as an occupational hazard:
Across repeat testing of treated cats, roughly one in four remained positive afterwards.6 Failure has several distinct causes, and they call for different responses:
| Cause | How you tell |
|---|---|
| Reinfection | Another untreated cat in the household; signs return after apparent clearance. |
| Under-dosing | Weight estimated rather than measured; missed doses; a lower dose used because of side-effect fear. |
| Poor compounding | Low drug content or altered release in a compounded product. Compounded formulations vary. |
| True resistance | Documented in vivo and in vitro, associated with increased oxygen-scavenging by the organism. Rare but real.7 |
| It was never this | The diagnosis rested on a smear, or on a PCR in a cat with no compatible signs, and something else is going on. |
This is the most important distinction on the site, and it is where most of what you will read online goes wrong. Web pages, and plenty of veterinary handouts, describe this infection as curable. That single word is doing a great deal of unearned work.
Eradication is achievable and has been documented: experimentally infected cats given the original high twice-daily dose were PCR-negative 21 to 30 weeks later.2 So "impossible to cure" would be wrong, and a page claiming that would be dismissed by any veterinarian who checked it.
But "curable" and "cured" are not the same claim, and only the second one matters to a household. What the evidence actually supports is narrower:
| State | How you know | What it means for other cats |
|---|---|---|
| Carrier | Positive on a properly collected test, with or without diarrhea. | Shedding. Can infect others. Symptom-free carriers are common and are the usual source of reinfection. |
| Suppressed | Diarrhea improved or resolved, but no retest, or a retest too soon. | Unknown. A normal stool is not evidence of clearance, and this is the state most treated cats are actually discharged in. |
| Cleared on retest | Negative PCR on a fecal loop sample, taken far enough after the last dose. | The only state that can be called a cure — and even then, only for that cat, at that time. |
Ronidazole exists for cats only as a compounded or repurposed product. It is not approved for any companion animal, it is compounded from chemical-grade powder by pharmacies, and it is also sold without prescription as a pigeon medication. Those routes are not equivalent, and the difference has been described in the literature as a cause of treatment failure:
So a cat that fails treatment is not automatically a resistant cat. It may have received a product that did not contain what the label said, in a form that did not deliver it. That is a quality problem in the supply chain, not a property of the cat.
The household consequence is not speculation, and it is not only owners saying it. The definitive review states directly that it is "a common misconception in multi-cat settings that only cats with diarrhea are infected", and that when presumed-uninfected cats are exposed to a cat that later fails treatment, "reinfection can never be ruled out".3
Which is the practical argument for the thing most households skip: do not consider a cat clear, and do not introduce it to a negative household, on the strength of a resolved stool or a single early negative test. Treat the cat, retest it properly, retest late enough to mean something, and treat the other cats as potentially positive until you have reason to think otherwise.
This list matters as much as the drug page, because most of it is still recommended somewhere.
| Agent | Verdict |
|---|---|
| Metronidazole | Stool often firms up while it is given and the diarrhea returns when it stops. The organism is not eradicated. Does not clear infection |
| Tinidazole | Reduces detection and confers some resistance to reinfection, but fails to eradicate in many cats. The fallback, and second-rate. Tested in cats |
| Secnidazole | The member of the family most often reached for after ronidazole fails, and the one with the least to go on. No published trial exists in cats with this infection. The feline dosing data comes from a different organism — a single 30 mg/kg dose studied in Giardia-infected cats12 — and the only T. foetus work is a 1983 biochemistry paper showing the organism reduces secnidazole to free radicals, which is a mechanism, not an outcome.13 It is not unreasonable to ask a veterinarian about; it is unreasonable to assume it is the gentler ronidazole. No published evidence |
| Fenbendazole, furazolidone | Fecal consistency may improve during treatment; the organism persists and diarrhea returns. Does not clear infection |
| Paromomycin | Do not. A published case series documented acute renal failure in four cats treated with paromomycin. Harm reported |
| Probiotics (Enterococcus faecium) | Inhibited the organism's growth, adhesion and cytotoxicity in the laboratory. No clinical trial in infected cats has been published. Prevention hypothesis only. Laboratory only |
| Colonic flushing as therapy | No published evidence. Flushing is a sampling technique; washing the colon out is not a treatment in any guideline. No published evidence |
A colon-targeted formulation of ronidazole in a guar-gum capsule achieved clearance in 84% of treated cats with no adverse effects at 30 mg/kg once daily8 — which would be the best result in the literature. It is not commercially available and, per the pharmacology literature, is not readily prepared by compounding pharmacies.
That is the single most interesting unanswered question in this field, and it is a manufacturing problem, not a biology problem. It is on the open questions page where it belongs.
