Proof of work

By the Blockhorizon Editorial Team · Updated 2 October 2026 · How we check facts

How does proof of work work?

The Bitcoin whitepaper describes it step by step. Nodes collect transactions into a block, then keep changing a number called a nonce and re-hashing the block until the hash begins with a required number of zero bits. Finding such a hash takes enormous trial and error. Checking it takes one calculation.

Whoever finds a valid hash broadcasts the block. Other nodes accept it only if every transaction in it is valid and not already spent, then start building the next block on top. Mining is the everyday name for this race.

Why is a proof-of-work chain hard to rewrite?

Each block is linked to the one before it, so changing an old block means redoing its work and the work of every block after it. Under the whitepaper's rule, the valid chain is the one with the most work invested, a system it calls one-CPU-one-vote.

The paper calculates that an attacker with 10% of the network's hash power has about a 0.09% chance of catching up from 5 blocks behind. That chance shrinks exponentially with each extra block.

How does Bitcoin keep blocks coming at a steady pace?

The puzzle's difficulty adjusts. NIST explains that Bitcoin resets it every 2016 blocks to aim for about one block every ten minutes. At that target pace, 2016 blocks take roughly 14 days, so the network recalibrates about twice a month.

What are the drawbacks of proof of work?

NIST names three costs: high computation, high power consumption and an arms race in specialised hardware. Those costs are the security, since attacking the chain means outspending honest participants, but they are also the main criticism.

Not every network uses it. Ethereum originally ran on proof of work and, according to ethereum.org, switched to proof of stake in September 2022.