Skip to content

Reorg (Chain Reorganization)

What a chain reorganization (reorg) is on Ethereum, why shallow one-block reorgs are routine while deep reorgs are a protocol concern, and how proposer boosting limits them.

A reorg (chain reorganization) happens when the network’s view of the canonical chain changes: a block that briefly sat at the tip gets displaced by a competing block, and every node following the rules switches to the new branch. It’s a normal side effect of a decentralized network propagating blocks over real, imperfect network latency — not evidence of an attack by itself.

Shallow, single-block reorgs happen on Ethereum from time to time when two validators produce blocks close together and the network briefly disagrees on which one arrived “first.” Ethereum’s consensus design includes a rule called proposer boosting, which gives extra weight to the block a validator actually attested to seeing, specifically to make these short forks resolve quickly rather than letting a late-arriving block with slightly more attestations win.

Why it matters

Depth is what distinguishes routine from alarming. A one-block reorg is noise. A multi-block reorg would mean something that was already treated as settled got rewritten, which is exactly what finality exists to make economically expensive. This is also why anything time-sensitive — large transfers, options expiries, bridge withdrawals — is usually built to wait for finality rather than a handful of confirmations.

Example

A validator with a slow or poorly configured node proposes a block a few hundred milliseconds late. The rest of the network has already built on an earlier sibling block, so the late block gets orphaned once the next block lands — a one-slot reorg that resolves itself without anyone needing to intervene.

Related · Protocol

← All glossary terms

Not financial advice.