What Is a 51% Attack? Blockchain Security Risk Explained

A 51% attack is when a single entity or coordinated group gains control of more than half of a blockchain's mining hashpower (in proof-of-work systems) or staked value (in proof-of-stake systems), allowing them to manipulate which transactions get confirmed and potentially reverse or double-spend recent transactions.

Blockchains rely on the assumption that no single party controls a majority of the network's block-production power. Consensus mechanisms are designed so that honest majority behavior keeps the chain secure; a 51% attack breaks that assumption directly.

How the attack actually works

In a proof-of-work chain, whoever controls the most computational power (hashrate) has the best odds of mining the next block and, cumulatively, of building the longest valid chain. Most nodes follow a rule that treats the longest valid chain as the authoritative one. If an attacker controls more than half the network's hashrate, they can, in principle, mine a private alternate chain faster than the honest network mines the public one.

The classic attack works like this: the attacker sends a transaction (for example, depositing crypto to an exchange) and gets it confirmed on the public chain. They then withdraw or trade using the resulting credit. Simultaneously, they've been secretly mining an alternate chain that does not include that transaction. Once their private chain is longer than the public one, they broadcast it. Nodes following the longest-chain rule switch over, effectively erasing the original transaction from consensus history — meaning the deposit never happened as far as the new canonical chain is concerned, even though the attacker already benefited from the exchange or counterparty acting on it. This is called double-spending.

An attacker with majority hashpower can also selectively censor certain transactions from being confirmed, though they cannot steal funds from addresses they don't control, alter the coin supply, or forge signatures — the attack only affects transaction ordering and confirmation, not the cryptographic security of individual wallets.

Which chains are actually vulnerable

Bitcoin and Ethereum are considered extremely resistant to 51% attacks in practice because acquiring a majority of their hashpower (Bitcoin) or staked ETH (Ethereum, since its move to proof-of-stake) would cost billions of dollars and would likely crash the value of the very asset the attacker is trying to profit from — a strong economic disincentive.

Smaller-cap chains are far more exposed, because renting sufficient hashpower or acquiring enough tokens is comparatively cheap relative to the value being attacked. Numerous smaller proof-of-work chains — including past incidents on chains like Ethereum Classic, Bitcoin Gold, and Verge — have suffered real 51% attacks resulting in confirmed double-spends against exchanges.

Relative attack cost by chain type

Chain characteristic	Attack cost	Real-world 51% attack risk
Large-cap PoW (e.g., Bitcoin)	Extremely high (billions in hardware/energy)	Very low
Small-cap PoW altcoins	Low to moderate (rentable hashpower markets exist)	Demonstrated multiple times
Large-cap PoS with broad stake distribution	Extremely high (requires majority of staked value)	Very low
Small PoS chains with concentrated validators	Lower, depending on validator concentration	Possible, less studied historically

Why proof-of-stake changes the calculus somewhat

In proof-of-stake systems, an attacker needs to control a majority of staked tokens rather than physical hardware. This introduces a different dynamic: acquiring that much stake typically requires buying a large share of the circulating supply on the open market, likely driving the price up as they buy, and most PoS chains also include "slashing" penalties that can destroy a validator's stake if malicious behavior is detected, adding a direct financial deterrent beyond just reputational cost. Our guide on proof-of-work vs proof-of-stake covers these mechanisms in more depth.

Practical implications for users

For everyday users, the main practical exposure to 51% attack risk comes when receiving payments in a smaller-cap coin — merchants and exchanges typically wait for a number of confirmations before treating a deposit as final, specifically to reduce the odds a shallow reorganization can reverse it. Waiting for more confirmations, and being cautious about which chains you rely on for high-value, time-sensitive transactions, reduces exposure. This is a different risk category from smart contract exploits or wallet-draining scams covered elsewhere in our common DeFi scams guide — a 51% attack is a consensus-layer risk, not a code-bug or social-engineering risk.

Bottom line

A 51% attack lets a majority-hashpower or majority-stake actor rewrite recent transaction history and double-spend funds, but it cannot steal from wallets it doesn't control or forge new coins out of thin air. Large, well-distributed chains like Bitcoin and Ethereum are effectively immune in practice due to the sheer cost involved, while smaller-cap chains with cheaper-to-acquire hashpower or concentrated stake remain realistically exposed. If you're transacting in a smaller-cap asset, treat confirmation counts and chain security budget as part of your risk assessment, not an afterthought.