Show MeDogs both eat rats and kill them, and the two behaviors belong to one sequence. A dog detects a rat, stalks or pounces, breaks the rat's spine with a shake, then may swallow part or all of the carcass. Killing is the more common outcome; eating is opportunistic and uneven across individuals. Which dog kills, which eats, and how dangerous either habit is depends on genetics, training, hunger, and the pathogen load the rat carries.
The distinction matters because owners conflate them. A terrier that kills 12 rats in a barn and leaves every carcass tells you nothing about whether it will eat a poisoned one. A retriever that has never killed anything may still gulp a dead rat it finds on a walk. The risks track the eating, not the killing.
The predation sequence, step by step
Canine predation on small mammals runs on a fixed motor pattern that ethologists call the predatory sequence. The components are orient, stalk, chase, grab, shake, kill, and consume. Each component is separately controlled in the brain, which is why selective breeding could amplify some steps and suppress others. Herding breeds retain stalk and chase but have weakened grab and kill. Terriers retain grab and shake with high intensity. Retrievers retain grab and carry but show weak kill and variable consume.
The kill itself is mechanical. A dog grabs the rat across the thorax or spine and shakes side to side. The acceleration forces at the rat's neck exceed what the cervical vertebrae and spinal cord can tolerate, producing fracture and cord disruption within a second or two. This is the same shake that lets a terrier dispatch a rat larger than its own head. It is not a bite-force problem; it is a whiplash problem.
Consumption sits at the end of the chain. It is gated by hunger, by the dog's history with food, and by whether the sequence has already been "completed." Working terriers often treat the kill as the terminal reinforcer and lose interest in the carcass. Dogs that have scavenged successfully, or that are fed once daily, are more likely to eat what they kill. The eating is not a separate instinct; it is the last link of the same chain, and it can be strengthened or weakened by reinforcement.
One behavioral caveat worth flagging: the sequence can run without a rat. A dog that shakes a stuffed toy or a squirrel it caught is performing the same motor program. Suppressing the shake in one context does not reliably suppress it in another.
How often dogs actually kill and eat rats
Prevalence is hard to pin down because most data come from working-dog surveys and shelter intake, not from controlled observation. What is clear from those sources is that the behavior is common enough to be a management issue and rare enough that most pet dogs never do it.
Working terrier breeds are the outlier. Ratting terriers in the UK and US are bred and trained specifically for the kill step, and a capable terrier on a barn hunt can account for dozens of rats in a single session. The kill rate per dog per hour is high in infested structures and near zero in clean ones; availability, not appetite, sets the ceiling.
Pet dogs are a different population. Veterinary behaviorists consistently report that killing small mammals is reported by a minority of owners, and that most kills are opportunistic encounters on walks or in yards rather than deliberate hunting. Eating the kill is reported in a smaller subset still. The strongest single predictor is prior scavenging behavior, not breed.
Ingestion numbers matter more than kill numbers for health risk. A single rat carcass weighs roughly 200 to 500 grams depending on species and age. That is a substantial protein and fat load for a small dog, and it can trigger pancreatitis in dogs with a history of the condition. Rats killed by anticoagulant rodenticides carry the toxin in liver and tissue, which is the pathway that makes carcass eating dangerous rather than merely unappetizing.
Comparison: kill behavior versus eat behavior
| Behavior | What triggers it | Breed skew | Health risk | Owner control |
|---|---|---|---|---|
| Stalk and chase | Movement, scent | Herding, hound | Low | Leash, recall |
| Grab and shake | Contact with prey | Terrier | Low to moderate (bite wounds) | Muzzle, avoid |
| Kill | Shake completion | Terrier, some hounds | Low to moderate (self-injury) | Prevent encounter |
| Eat carcass | Hunger, scavenging history | Weak breed skew | High (toxins, pathogens) | Leash, "leave it" |
| Roll on carcass | Scent marking | Any | Low (odor, not disease) | Bathe, manage |
| Dig for rats | Burrow scent | Terrier, dachshund | Moderate (injury, toxin contact) | Fence, supervise |
The table's point is that the risk column does not follow the breed column. The behaviors with the strongest genetic skew are the least dangerous to the dog; the behavior with the weakest skew, carcass eating, carries the highest health risk. Owners who focus on suppressing the kill are often managing the wrong link.
The toxin pathway: why eating is the real hazard
Anticoagulant rodenticides, chiefly the second-generation compounds, work by blocking vitamin K recycling in the liver. The rat dies of internal hemorrhage over several days. The toxin remains in the rat's liver at concentrations far above what killed it, because the rat cannot metabolize it efficiently. A dog that eats that liver gets a dose proportional to the liver's burden, not to the rat's body weight, which is why a small amount of liver can be clinically significant.
The clinical picture in dogs is delayed. Signs of anticoagulant toxicosis typically appear 3 to 7 days after ingestion: lethargy, pale gums, weakness, and bleeding from the nose, gums, or GI tract. Clotting times (PT, aPTT) are prolonged before signs appear, so a dog that ate a poisoned rat 24 hours ago can look normal and still be in trouble. Veterinary management involves vitamin K1 therapy and monitoring, and the decision to treat is made by the veterinarian based on exposure history and clotting tests, not by the owner.
The other toxin pathway is secondary poisoning from rats that died of other causes. Rats are reservoirs for leptospirosis, and canine leptospirosis is a recognized risk in dogs exposed to rat urine and tissue, particularly in areas with standing water. Salmonella and other enteric bacteria are also carried by rats and can colonize a dog that eats a carcass. These are not theoretical: public health agencies treat rat-dog contact as a leptospirosis exposure route.
A counterintuitive detail: the rat a dog kills itself is often safer than the rat it finds dead. A live rat the dog just dispatched has not had time to accumulate a lethal rodenticide burden, and its tissues are fresh. A carcass found on a walk has an unknown cause of death and an unknown toxin load. The instinct to eat what you kill is less dangerous than the scavenging that follows a found carcass, which is why leash control matters more than kill suppression.
Timeline: what happens after a dog ingests a rat
| Time after ingestion | What is happening | What an owner may see |
|---|---|---|
| 0 to 2 hours | Stomach acid begins breaking down tissue; bacteria load enters gut | Nothing, or mild GI upset |
| 2 to 12 hours | Bacterial translocation; toxin absorbed if rodenticide present | Vomiting, loose stool in some dogs |
| 12 to 24 hours | Coagulation factors begin to deplete if anticoagulant present | Still often normal |
| 1 to 3 days | Clotting time prolonged before signs appear | Lethargy, reduced appetite |
| 3 to 7 days | Clinical bleeding if toxicosis is developing | Pale gums, nosebleeds, bloody stool |
| 7 to 14 days | Recovery if treated; relapse if vitamin K stopped early | Improving or worsening |
| 14 to 30 days | Full resolution in uncomplicated cases | Normal |
The table's practical value: the dangerous window for anticoagulant toxicosis is days 3 to 7, not the first hour. Owners who relax after 24 hours are relaxing at the wrong time. Any dog with a known rat-carcass ingestion should have a veterinary conversation, and clotting tests are the way to settle the question.
Red flags that change the response
Dog ate a rat found dead on a walk or in a yard, especially near bait stations or agricultural buildings. Dog shows pale gums, nosebleeds, bloody stool, or unexplained bruising at any point in the 14 days after ingestion. Dog vomited blood or passed black tarry stool. Dog has a history of pancreatitis and ate a full carcass rather than a fragment. Dog is unvaccinated against leptospirosis and had contact with rat urine or standing water. Dog is small-bodied (under 10 kg) and ate a large rat, increasing the relative toxin dose.
These are triage signals, not diagnoses. The presence of any one warrants a call to a veterinarian rather than a wait-and-see approach.
Practical actions and thresholds
The goal is not to eliminate predation, which is largely hardwired, but to manage the eating step and the encounter step.
Keep the dog leashed or under voice control in areas with known rat activity. The single most effective intervention is preventing the dog from reaching a found carcass before it can be swallowed. A reliable "leave it" cue and a high-value treat reward for compliance are worth more than any breed-specific training.
Do not use rodenticides in areas the dog can access, or use them only in tamper-resistant bait stations placed where the dog cannot reach the carcass. A dog that kills a rat in a treated building has a real exposure risk; a dog that kills a rat in an untreated barn does not. Ask before you bait.
Check the dog after any known kill or carcass contact. Inspect the mouth for bites and the gums for color. If the dog ate part of the rat, call the veterinarian with the dog's weight and the estimated amount eaten. Do not induce vomiting unless instructed.
Vaccinate for leptospirosis in areas with rat activity or standing water. This is a preventive measure, not a treatment, and its value depends on local risk.
Manage the environment. Sealing entry points, removing food waste, and reducing harborage lowers rat density, which lowers encounter frequency. A dog cannot eat a rat it never meets.
For dogs with high prey drive and a history of eating kills, consider a basket muzzle during off-leash time in high-risk areas. It prevents ingestion without preventing exercise.
Thresholds to reconcile: the 3-to-7-day window for anticoagulant signs, the roughly 200-to-500-gram rat carcass size, and the under-10-kg body weight flag all point to the same conclusion. Small dogs that eat whole rats in treated areas are the highest-risk combination, and they should be evaluated rather than observed.
What the evidence cannot settle
The behavioral literature on dog-rat predation is thin and skewed toward working terriers, so prevalence estimates in pet dogs are unreliable. There is no large-scale study that quantifies how often pet dogs eat what they kill, and no validated tool that predicts which individual will eat a carcass. Breed stereotypes are weak predictors of the eating step; prior scavenging and food motivation are better, but neither is a clean test.
The toxicology is better established but still inference-heavy for secondary exposure. Dose-response for a dog eating a rodenticide-poisoned rat depends on the specific compound, the rat's liver burden, and the dog's size and clotting status, and those variables are rarely known in a real exposure. Veterinarians make treatment decisions on exposure history and clotting tests, not on a published conversion factor, because no reliable one exists.
What can be concluded is narrower and more useful. Dogs kill rats as part of a fixed predatory sequence, and they eat them as the sequence's final, variable step. The killing is mostly genetic and mostly safe for the dog. The eating is mostly learned and is the step that carries the toxin and pathogen risk. Managing the encounter and the ingestion is more effective than trying to breed or train out the instinct. And the clock that matters after ingestion runs in days, not hours, which is why the first 24 hours of apparent normality are not reassurance.