Table · StrataPlanned

Biggest Bridge Hacks, Ranked by Losses and What Failed

Which cross-chain bridges have been hacked for the most, and what failed in each: keys, verification or code?

Cross-chain bridge exploits ranked by value taken, with each bridge's design, the failure behind the loss and how much came back.

ShareEmail itvoltfincrypto.com/apps/biggest-bridge-hacks
Blueprint

What the table will hold.

The table as it will open in Strata. Each figure is named here and arrives with its receipt; until its data is published, its place stays empty.

One row One row per bridge exploit, valued at the attack blockSorted by Value taken, highest first

RankBridgeDateChainsDesignSigner thresholdWhat failedLoss typeValue takenUSDRecoveredUSDAttributionReceipt
Rank rank Number
Position by value taken at the time of the attack.
Bridge bridge Text
The bridge and the version of it that was attacked.
Date date Date
The day of the first attack transaction.
Chains route Text
The chains the bridge connected where funds were taken or minted.
Design design Text
How the bridge verified messages: a multisig, an off-chain key set, a light client, an optimistic scheme or external verifiers.
Signer threshold signer_threshold Text
Signers or validators required to approve a message at the attack block, such as five of nine.
What failed failure Text
Compromised keys, a verification flaw, a contract bug or a misconfiguration, as the post-mortem found.
Loss type loss_kind Text
Escrow drained, unbacked mint or both.
Value taken value_taken Exact amount
Token amounts taken or minted in the attack transactions, valued at the attack block.
Recovered recovered Exact amount
Value returned, frozen or seized from the attacker, valued at the attack block.
Attribution attribution Text
The group publicly tied to the attack, where an authority or the bridge team has named one.
Receipt receipt Link
The attack transactions, blocks and escrow reads behind the row.

Filter by Design, What failed, Loss type.

About

What it answers, and for whom.

Bridges hold pools of tokens on one chain to back copies on another, and they trust some mechanism to report what happened on the other side. When that mechanism fails, the loss is the whole escrow or an unlimited mint. This table ranks every recorded bridge exploit by value taken at the time and names what failed: keys, a verification check, contract code or the configuration around it.

The pattern repeats. Ronin's validator keys were taken in 2022, Wormhole's signature check on Solana was bypassed the same year, a faulty initialization let anyone forge Nomad messages and the KelpDAO rsETH bridge on LayerZero was exploited in April 2026. Each row records the bridge's design and its signer threshold at the attack block, so incidents can be compared by cause and not only by size.

Values are the token amounts that left escrow or were minted without backing in the attack transactions, priced at the attack block. Recoveries sit beside each loss, so the table also shows how much came back.

Method

How it will be measured.

Every figure will be read at a block, kept with its source and its exact read, and open a receipt anyone can re-check.

How receipts work →
  1. An incident enters the table when funds left a bridge's escrow, or tokens it never backed were minted, in attack transactions identified on-chain and confirmed by the bridge team or a public post-mortem.
  2. Value taken is the token amounts in the attack transactions, valued at oracle prices at the attack block; escrow drains and unbacked mints are labelled, since a mint can exceed what the escrow held.
  3. The bridge's design and its signer threshold at the attack block are read from its contracts where they are on-chain, and from its documentation otherwise, with the source recorded.
  4. Failure types follow the post-mortems: compromised keys, a verification flaw, a contract bug or a misconfiguration, each with the source that established it.
  5. Recovered value counts funds returned, frozen or seized from the attacker, valued at the attack block, so the share recovered is not distorted by later price moves.

The data it needs

Chain data, with some off-chain sources

  • security_incidentsTo buildHacks and exploitsA registry of exploits, hacks and rug pulls: date, chain, protocol, attack type, amount lost and the transactions involved.
  • bridge_flows_dailyTo buildBridge deposits and withdrawalsDeposits into and withdrawals out of canonical and major third-party bridges by chain, token and day, plus escrow balances at day end.
  • native_prices_dailyTo buildOracle prices by dayReference prices of major assets read from on-chain oracle feeds (Chainlink latestRoundData) at each day's last block, with each round's update time.
  • address_labelsTo buildAddress labelsLabels for known addresses: exchanges, bridges, protocols, funds, treasuries, sanctioned addresses and known attackers.
  • Attack transactions: token transfers out of bridge escrows and unbacked mint events on destination chains
  • Escrow balances at the blocks before and after each attack
  • Bridge signer sets and thresholds read from the bridge contracts at the attack block, where they are on-chain
  • Post-mortems and public disclosures by the bridge teams, recorded with their source (off-chain)
Questions

Questions people ask.

What is the biggest bridge hack?

Measured by value at the time, the Ronin bridge breach of March 2022 is the largest bridge hack widely reported: attackers controlled enough validator keys to sign withdrawals of ETH and USDC out of the bridge. The table ranks every bridge incident by the same measure, so later incidents take their place in order.

Why are bridges hacked so often?

A bridge concentrates value: its escrow backs every copy of a token on other chains. It also has to trust a report of events on another chain, through signers, a light client or a set of verifiers, and that trust is the usual point of failure. One flaw can empty the escrow or mint copies without limit.

What is the difference between an escrow drain and an unbacked mint?

In an escrow drain the attacker withdraws the tokens a bridge holds on its home chain. In an unbacked mint the attacker makes the bridge issue copies on another chain without any deposit, then sells or borrows against them. A mint leaves holders of the copies with tokens nothing backs.

Which bridge designs have failed?

Bridges secured by a small set of keys fail when those keys are stolen or misused, as at Ronin and the Harmony bridge in 2022. Bridges that verify proofs fail through flaws in the verifier, as at Wormhole and the BNB Chain token hub. The table records each bridge's design so the pattern can be checked, not assumed.

How much of a bridge hack is usually recovered?

It varies widely. Some attackers returned most of the funds after negotiation, some freezes caught part of the loot and some losses were never recovered at all. Each row shows its own recovered value, and the stolen crypto recovered table breaks recoveries down by how they happened.

Escrow
The contract that holds tokens on their home chain while copies circulate on other chains; also called a lockbox or vault.
Unbacked mint
Tokens a bridge issues on a destination chain without a matching deposit on the source chain, usually after a forged message.
Signer threshold
How many of a bridge's signers or validators must approve a message, such as five of nine; the fewer keys needed, the fewer an attacker must steal.

Build Biggest Bridge Hacks, Ranked by Losses and What Failed on verified data.

Start free in Voltfin Studio: boards, tables and reports where every number opens its receipt. Tell us which data you need first and we will build it next.