Corvid intelligence
Corvids are birds in the family Corvidae, which includes crows and ravens (genus Corvus), along with jays, magpies, nutcrackers, jackdaws, rooks, and choughs — well over a hundred species spread across most of the world. Several of these birds have been studied closely in comparative cognition, and a handful of their behaviors are documented through controlled experiments and long-term field work rather than anecdote alone. This page gives a broad, cautious overview of four areas researchers return to: flexible problem-solving, food caching, social learning, and tool use in particular species.
The examples below are representative, not a complete catalogue, and they are not evenly distributed across the family. A finding in one species — say, a New Caledonian crow in an aviary or a Clark's nutcracker recovering seed stores — does not automatically describe all corvids. The aim here is to describe what specific studies on specific species actually show, to flag whether the evidence is from captivity or the wild, and to avoid both "smartest bird" rankings and the assumption that every corvid does what the most famous one does.
A representative, source-cautious overview of how corvids — crows, ravens, and jays — solve problems, cache food, learn socially, and, in some species, use tools, without ranking them or treating one famous species as the whole family.
Representative, not complete:
These examples are representative of particular corvid species and study populations, not a complete or family-wide account. Corvidae contains huge diversity — crows, ravens, jays, magpies, nutcrackers, jackdaws, and more — and an ability documented in one species (such as New Caledonian crow tool making or Clark's nutcracker seed memory) should not be assumed for all corvids. This page describes what specific studies show, flags captive versus wild evidence, and offers no intelligence ranking or score.
Representative behavior themes
- Flexible problem-solving in puzzle and apparatus tasksEvidence: Controlled study
Several corvids solve multi-step puzzles in controlled settings. Corvus species have dropped sinking objects into a water-filled tube to raise a floating reward in the Aesop's-fable paradigm, common ravens (Corvus corax) have performed coherent string-pulling to haul up hung food, and New Caledonian crows (Corvus moneduloides) have completed sequential tasks requiring actions in a set order. These results suggest attention to a problem's functional structure rather than pure trial-and-error within one attempt, but performance varies widely between individuals, and whether success reflects causal understanding or learning across repeated trials is genuinely debated in the literature.
- Food caching and spatial memoryEvidence: Mixed evidence
Many corvids store surplus food in scattered caches and recover it later, and some are specialists. Clark's nutcracker (Nucifraga columbiana) buries large numbers of pine seeds and relies heavily on spatial memory to relocate them across a season, and scrub-jays (genus Aphelocoma) have been studied for how they manage perishable versus durable stores. Caching ability and the precision of recovery differ a great deal between species, so this is a representative corvid behavior, not a uniform one, and recovery is better described as ecological memory in context than as a flawless map.
- Social learning and observation around cachesEvidence: Mixed evidence
Some corvids attend closely to what other birds do, especially around food. Aviary studies of common ravens report that birds which watched another cache could later relocate those stores using observational spatial memory, and that cachers adjusted their hiding — moving or delaying — when a competitor could see them. Field experiments on American crows (Corvus brachyrhynchos) indicate they can learn and remember a human face tied to a threatening event and that the response can spread among local birds. Whether such behavior reflects tracking another animal's visual access, or something closer to understanding its mind, is contested; simpler explanations attending to a competitor's behavior are not fully ruled out.
- Tool use concentrated in a few species, not the whole familyEvidence: Mixed evidence
Habitual, skilled tool use is the exception among corvids, not the rule. The New Caledonian crow (Corvus moneduloides) is the standout: in the wild and in captivity it uses sticks and modified plant material to extract prey and manufactures hooked and stepped pandanus-leaf tools. The Hawaiian crow, or alala (Corvus hawaiiensis), has also been reported using sticks to forage. Most crows, ravens, and jays are not habitual tool makers, so it is inaccurate to treat New Caledonian crow tool use as a general corvid trait, and even within tool-using species the balance of inherited predisposition versus learning is still studied.
Corvids are a diverse family, not one famous bird
"Corvid intelligence" is often illustrated with a single dramatic example — a New Caledonian crow shaping a hooked tool, or a raven solving a string puzzle — but Corvidae is a large and varied family, and no one species stands in for all of it. Crows and ravens (genus Corvus), jays, magpies, nutcrackers, jackdaws, rooks, and choughs differ in ecology, social structure, and the kinds of problems they routinely face, and the behaviors studied in one are not automatically present in another.
This matters because the most striking corvid findings come from a small set of intensively studied species and, often, from captive or aviary work. Long-running research programmes have concentrated on a few birds — common ravens, several Corvus crows, New Caledonian crows, scrub-jays, and Clark's nutcracker among them — so the published picture is detailed for those species and much sparser for the rest of the family. Describing the group honestly means naming which species an observation comes from rather than attributing it to "corvids" as a whole.
How the evidence is gathered, and its limits
Corvid cognition is studied through controlled experiments, aviary observation, and field work, and each approach has trade-offs. Apparatus tasks — water tubes, string-pulling, sequential puzzle boxes — let researchers control variables, but they test captive birds in artificial situations and may not reflect everyday wild behavior. Field experiments, such as those on American crow face recognition, observe wild birds but rely on indirect measures like scolding and mobbing responses.
A recurring theme in the literature is caution about interpretation. A clever-looking result can sometimes arise from associative learning across repeated trials, attention to a companion's behavior, or subtle cues, rather than from reasoning or understanding. Careful researchers build in controls, report how much individuals vary, and flag where claims are contested — for example, whether water-displacement performance shows causal understanding, or whether cache-protection reflects tracking what a competitor saw versus a human-like belief about its mind. This page follows that posture: it separates what is observed from what is inferred and notes where the science is unsettled.
What this page does not claim
That corvids are the "smartest" birds or animals, or that they can be given any intelligence score or ranking — comparative cognition does not support such a single scale.
That all corvids use tools; habitual, skilled tool making is documented mainly in the New Caledonian crow, with some reports in the Hawaiian crow, not in crows, ravens, or jays generally.
That solving puzzles or string-pulling proves human-style reasoning or insight; the interpretation is debated, much evidence is captive, and individual performance varies widely.
That corvid social learning, cache-protection, or face recognition demonstrates human-like beliefs, theory of mind, or moral judgment of people — these are evidence-bounded claims about behavior, not proof of inner experience.
Any handling, attracting, feeding, training, hunting, pest-control, or wildlife-approach guidance; this is an educational ethology overview, not how-to advice.
Related animal profiles & behavior pages
How these claims are studied
Group-level behaviour is easy to overstate, so these claims are kept cautious and labelled by evidence context. See how animal intelligence is studied, why animal iq rankings mislead, and animal research sources for our methodology.
Frequently asked questions
- Are corvids the smartest birds?
- There is no scientifically meaningful "smartest bird" and this page gives no ranking. Several corvids perform well on particular cognition tasks, but different species are good at different things, and a test that suits one bird can be almost meaningless for another. It is more accurate to describe specific, documented abilities in specific species — such as Clark's nutcracker spatial memory or New Caledonian crow tool use — than to crown a winner.
- Do all crows, ravens, and jays use tools?
- No. Habitual, skilled tool making is documented mainly in the New Caledonian crow (_Corvus moneduloides_), with some tool use reported in the Hawaiian crow (_Corvus hawaiiensis_). Most crows, ravens, and jays are not habitual tool users, so New Caledonian crow tool use should not be treated as a general corvid trait.
- How do corvids remember where they hid food?
- Caching corvids rely heavily on spatial memory. Clark's nutcracker, for example, buries many seeds and recovers them across a season, and ravens that watched another bird cache can relocate those stores. Abilities vary widely between species, and recovery is best understood as memory used in an ecological and social context, not as a flawless internal map.
- Does corvid problem-solving prove they think like people?
- It does not. Corvids succeed at several controlled puzzles, but researchers debate what the successes mean — some results once read as causal understanding may also come from learning across repeated trials, and individuals differ a lot. Their behavior is best described as flexible problem-solving in specific tasks, not as evidence of human-style reasoning or inner experience.
Last updated: