Skip to content
NatureHQ

Claim check

Do ravens hold grudges?

Exaggerated

Something true sits underneath, stretched well past what was shown.

The memory is real and unusually long. The grudge is not something the experiments can see. Ravens recognise former companions by voice after up to three years, and their response differs depending on whether that individual was an ally or a rival — which is a statement about memory, not about intent to retaliate.

The claim as it circulates

“Ravens remember people and birds who wronged them and will take revenge years later.”

Where you may have met it: Social media posts about corvid revenge; Popular science books on bird intelligence; Coverage that merges raven and crow findings into one bird

Taking it apart

What was claimed
That ravens remember individuals who treated them badly and act against them later, in a way that amounts to holding a grudge.
What was actually observed
Captive ravens were played recordings of birds they had lived with years earlier, some former allies and some former rivals, alongside calls from strangers. Their vocal responses differed by category, up to three years after the birds had last met. In a separate design, ravens responded more strongly to playback implying that two other birds had swapped rank — including birds in a neighbouring group they were not part of.
What the evidence supports
That ravens retain socially differentiated memories of specific individuals for years, and that they track relationships between third parties rather than only their own.
What it does not support
That a raven intends to retaliate, or would act against a remembered rival given the chance. The measure is a vocal response to a recording in captivity. Nothing in these designs tests behaviour towards the individual in a real encounter, and no study has shown a raven acting against a human who mistreated it.

Two things get merged in the popular version. The crow face-recognition work shows wild birds responding to a specific human face associated with being caught — that is about humans, and it is about alarm. The raven work is about other ravens, and it is about memory of a relationship. Neither shows retaliation, and they are frequently reported as though they were one finding about one bird.

The genuinely striking part is the third-party tracking. Noticing that two other individuals have changed places in a hierarchy requires holding a relationship in mind that you are not part of, and doing it for a group you merely live alongside means there is no direct stake in the outcome. That is a more interesting claim than revenge, and it is better evidenced.

Where the evidence lives

The claims underneath

Each one carries its own evidence, scope and caveats. Expand any of them to reach the studies.

Ravens remember individual birds, and whether they got on with them, for years

Well supported

Good evidence backs this, though some details remain open.

Common ravens (Corvus corax) responded differently to playback of calls from former group members than from strangers up to three years after separation, and distinguished former affiliates from former rivals.

Who this applies to
captive common ravens in managed social groups
Studied in
Corvus corax
Why we rate it this way, and what the caveats are
Well supportedModerate confidence

A clear playback result with an appropriate stranger control. Confidence is moderate because vocal response is a narrow window onto memory, and the birds were captive with researcher-arranged social groups.

Caveats

  • This is the serious core of the popular "ravens hold grudges" claim — and much narrower than it.
  • A grudge implies intent to act; nothing here measures that.
  • Captive birds, whose social history was arranged rather than chosen.

Still unanswered

  • Would remembered relationships change behaviour in a real encounter?

Last reviewed 2026-08-09

The evidence (1 study)

Ravens notice when two other birds swap rank — even in a group they are not in

Emerging evidence

Real findings exist, but too few or too recent to be settled.

Common ravens responded more strongly to playback sequences implying a dominance reversal between two other individuals than to sequences consistent with the existing hierarchy, including for birds in a neighbouring group.

Who this applies to
captive ravens in established social groups
Studied in
Corvus corax
Why we rate it this way, and what the caveats are
Emerging evidenceModerate confidence

A well-designed playback study with an appropriate consistent-hierarchy control. Held at emerging because the result rests on constructed audio events and has not been widely replicated.

Caveats

  • Playback constructs an event that never happened, which birds may find odd for other reasons.
  • Captive groups have unusually stable and well-known hierarchies.
  • Whether ravens represent relationships or track reliable behavioural cues is unsettled.

Still unanswered

  • Does this hold in wild raven populations with fluid membership?

Last reviewed 2026-08-09

The evidence (2 studies)

Go deeper

Common raven

They hide food more carefully when they might be watched — with nobody in sight.

Last reviewed 2026-08-09