Common claims about pigeons
Last reviewed 23 August 2026. Five things said about feral pigeons often enough to be worth setting the published evidence against. One of them is true, one is unanswerable on current data, and the rest are more qualified than either side tends to allow.
“They are vermin”
There is no category in UK law called vermin that a bird can be placed in. The feral pigeon is a wild bird under the Wildlife and Countryside Act 1981, which makes killing or taking one an offence except under licence — the same Act, and the same offence, that protects any other wild bird. What a general licence permits, and to whom, is set out on the law and evidence page.
On the ornithological side the classification is equally specific. The British Ornithologists’ Union places Columba livia in categories A and C4, and the BTO records it as Green listed on Birds of Conservation Concern and of Least Concern in Europe.
None of that answers the thing people usually mean by the word, which is that a flock fouls a building and costs money to manage. That is a real problem and a procurement question rather than a classification one; the methods, their costs and their evidence base are on the non-lethal alternatives page.
“They are not really wild birds”
This one is correct, and it is the more interesting fact of the two. The BOU’s definition of category C4 is “naturalized feral species – species which have been domesticated which now have populations established in the wild”, and the example the BOU gives for the category is the pigeon itself. The BTO redirects anyone searching its species database for Feral Pigeon to its Rock Dove page, on the stated ground that the rock dove “is the ancestor of the Feral Pigeon”.
The genetics are being worked out on the wild side of that relationship rather than the feral one: Hernández-Alonso and colleagues sequenced whole genomes from 65 historical rock doves spanning the species’ original range, including three specimens from Darwin’s collection (Molecular Biology and Evolution, 2023, open access).
The domestic history is documented rather than inferred. The PDSA Dickin Medal, instituted in 1943, has been awarded 75 times; 32 of those went to pigeons, the second largest group after dogs. Racing birds are still ringed and still traced through the Royal Pigeon Racing Association, which is why a bird with a “GB” ring is treated differently from one without — see found a pigeon.
“There are more of them every year”
Nobody can currently say. The BTO gives a UK breeding population of 465,000 pairs, dated 2016, for Columba livia as a whole, and states in the same entry that the species is poorly monitored and that the long-term Breeding Bird Survey trend shows a moderate decline that is not statistically significant. Pure rock doves are thought to be confined to remote coastlines in northern and western Scotland and in Northern Ireland; feral pigeons are widespread across the rest of the UK and absent from many upland areas.
The other half of the arithmetic is missing too. No public body is required to record or report lethal control carried out under a general licence, so the number of birds killed each year in the UK is not counted anywhere. The only figures that exist have come out through Freedom of Information requests, one authority at a time, and the law and evidence page carries a template for making one.
Source: BTO BirdFacts, Rock Dove, citing Balmer et al. (2013) for distribution.
“They spread disease”
Partly. The standard survey is Haag-Wackernagel and Moch, Health hazards posed by feral pigeons (Journal of Infection, 2004), which searched the epidemiological literature for reports of transmission from feral pigeons to people. It found 176 documented transmissions between 1941 and 2003. Feral pigeons harboured 60 different organisms pathogenic to humans, but only seven were ever transmitted, most commonly Chlamydophila psittaci and Cryptococcus neoformans. Aerosol accounted for 99.4% of incidents — that is, dried droppings disturbed into the air rather than contact with a bird.
Both halves of that paper get quoted, rarely by the same person
Its conclusion on the general population: although feral pigeons pose sporadic health risks, “the risk is very low, even for humans involved in occupations that bring them into close contact with nesting sites”.
Its conclusion on everyone else: “In sharp contrast, the immunocompromised patient may have a nearly 1000-fold greater risk of acquiring mycotic disease from feral pigeons and their excreta than does the general population.”
The first sentence is the one campaigners quote and the second is the one pest control quotes. The paper says both, and the distinction it draws is between populations, not between doses.
On avian influenza the RSPCA states that pigeons are highly resistant to it, while also saying that a bird showing signs of it should not be picked up or transported by a member of the public. What those signs are, and who to call, is on found a pigeon.
“They are stupid”
The experimental literature says otherwise, and it is worth quoting carefully, because on this subject the overclaiming runs in the pigeon’s favour and a citation that turns out to be overstated costs more than it gains.
- Medical images. Levenson, Krupinski, Navarro and Wasserman trained pigeons to distinguish benign from malignant human breast histopathology, and the birds generalised to image sets they had not seen before; they also detected cancer-relevant microcalcifications on mammograms. On a third task — classifying suspicious mammographic masses, which is difficult for people too — the authors record that the birds “proved to be capable only of image memorization and were unable to successfully generalize” (PLOS ONE, 2015, open access).
- General intelligence. Flaim and Blaisdell ran a battery across associative learning, memory, cognitive flexibility and reaction time. With the reaction-time task included they found a two-component structure influenced by the birds’ age; with it excluded, a component with the properties of the g factor found in humans (Animal Cognition, 2024, open access).
- Self-recognition. This is the claim most often overstated. Toda and Watanabe trained five pigeons to respond to live video of themselves and not to footage recorded earlier, and the birds learned to tell the two apart. The authors attribute the result to the temporal contiguity between the bird’s own behaviour and the visual feedback, and describe the method as “an important step toward improving the experimental analysis of self-recognition in non-human animals” (Animal Cognition, 2008). It is not a mirror self-recognition result, and the paper does not present it as one.
- Against primates. Scarf and Colombo tested pigeons and primates on numerical competence and orthographic processing and found comparable cognitive signatures — no qualitative difference in mechanism — while the training time required “revealed a clear quantitative difference” (Frontiers in Psychology, 2019, open access).
Every source above, with each one labelled for what it is, is on the sources page. What the law permits and what the published evidence says about managing numbers is on the law and evidence page.