How Does a Blockchain Work?

A blockchain is a ledger of transactions that thousands of computers keep in sync without trusting each other, using two tools: cryptographic hashes that make any edit to history instantly detectable, and a consensus process that decides whose version of new transactions everyone accepts. The result is a database nobody owns but everybody can verify — and that nobody can quietly rewrite.

This guide explains the machinery without math. You only need three ideas: hashes, blocks, and consensus.

Hashes: tamper-evident fingerprints

A hash function takes any input — a word, a file, an entire ledger — and produces a short, fixed-length fingerprint. Bitcoin's hash function, SHA-256, always outputs 64 hexadecimal characters, whatever you feed it.

Two properties make hashes the foundation of everything else:

1. The same input always gives the same fingerprint. Anyone can recompute it and check.
2. Any change to the input — even one character — produces a completely different, unpredictable fingerprint. Hashing "Alice pays Bob 5" versus "Alice pays Bob 6" gives two outputs with no visible relationship.

Crucially, you cannot work backwards from a fingerprint to the input, and you cannot craft an input that produces a fingerprint you've chosen in advance. A hash is a commitment: publish the fingerprint, and you're locked to exactly that data.

If a bank published the hash of its ledger every day, customers could detect edits — but they'd still have to trust the bank about which ledger is real. Blockchains close that gap with the next two ideas.

Blocks: batching transactions and chaining them together

Transactions on the network aren't recorded one at a time. They're collected into blocks — batches of typically a few hundred to a few thousand transactions — added to the ledger at intervals (roughly every 10 minutes on Bitcoin, roughly every 12 seconds on Ethereum).

Each block contains three things that matter here:

• The batch of transactions.
• A timestamp and some bookkeeping.
• The hash of the previous block.

That last item is the "chain" in blockchain. Block 900,001 contains the fingerprint of block 900,000, which contains the fingerprint of block 899,999, and so on back to the very first block in 2009. History is not a pile of independent records; it is a single linked sequence where every block vouches for everything before it.

Why this makes editing history loud

Suppose an attacker wants to alter a transaction in a block from last month. Changing even one character changes that block's hash. But the next block recorded the old hash — so the next block no longer matches and must be rewritten too, which changes its hash, which breaks the block after that. One edit forces rewriting every subsequent block, in public, while the honest network keeps extending the real chain.

A blockchain doesn't make tampering impossible; it makes tampering unhideable and enormously expensive. That distinction is the honest way to describe its security.

Consensus: agreeing without a referee

Hashing protects existing history. The harder problem is the present: thousands of computers, no manager, and everyone must agree on which new block comes next. Otherwise someone could spend the same coin twice by telling different halves of the network different stories — the "double-spend" problem that stumped digital-cash designers for decades.

Blockchains solve it by making the right to add a block costly, so an attacker can't cheaply flood the network with a false version of events. The two dominant designs differ in what the cost is.

Proof of work: cost = electricity

On Bitcoin, computers called miners compete to add each block by brute-force searching for a rare hash — recomputing the block's fingerprint trillions of times with tiny variations until one lands below a target. Finding it takes enormous computation; verifying someone else found it takes a millisecond. The winner broadcasts the block and earns newly issued coins plus fees.

The rule everyone follows: the valid chain with the most total work wins. To rewrite last week, an attacker must redo last week's mining and outpace all honest miners combined, sustained, which for Bitcoin means out-computing a global industry burning real electricity. Nobody has ever rewritten meaningful Bitcoin history this way.

Proof of stake: cost = capital at risk

On Ethereum and most newer chains, validators put up a deposit of the network's own coin (32 ETH to run an Ethereum validator). The protocol pseudo-randomly selects who proposes each block, and the rest attest to it. Misbehave — sign two conflicting histories, for example — and the protocol destroys part of your deposit, a penalty called slashing.

Same idea, different currency of cost: attacking a proof-of-work chain wastes electricity; attacking a proof-of-stake chain burns your own money. Proof of stake uses a tiny fraction of the energy, which is why Ethereum switched to it in 2022, at the cost of a more intricate protocol.

Following one transaction through the system

Put the pieces together by tracing a payment:

1. You sign. Your wallet builds a message — "send 0.1 ETH from my address to Maria's" — and signs it with your private key. The signature proves the owner authorized it, without revealing the key.
2. You broadcast. The transaction goes to the network and sits in a public waiting room (the mempool) with thousands of others.
3. A block includes it. A miner or validator packages your transaction into a candidate block, wins the right to publish, and broadcasts it.
4. Everyone verifies. Every node independently checks every transaction in the block — valid signature? sufficient balance? rules followed? An invalid block is simply rejected no matter who produced it. This is the quiet superpower: verification is cheap and universal, so producers of blocks have no authority, only a temporary turn.
5. The chain grows past it. Each subsequent block buries yours deeper. After a handful of blocks, rewriting history to remove your transaction would require redoing all of them against the whole network — which is why merchants wait for a few "confirmations" before treating large payments as settled.

At no point did any single party decide the truth. The ledger is whatever the rules, applied by everyone, agree it is.

What blockchains don't do

Beginner materials oversell this technology, so calibrate:

• "Immutable" has fine print. Deep history is effectively fixed, but the last few minutes can reorganize slightly as the network converges — hence confirmations.
• Garbage in, garbage out. The chain guarantees the ledger's integrity, not the honesty of what people write to it. A scam transaction is recorded just as faithfully as a legitimate one.
• The chain can't protect your keys. Nearly all crypto losses come from stolen or lost keys and social engineering, not broken consensus.
• Decentralization is a spectrum, not a checkbox. A chain validated by hundreds of thousands of independent nodes and one validated by a dozen server operators are both "blockchains." The security story is only as good as how distributed the participants really are.

Key Takeaways

• Hashes give every block a tamper-evident fingerprint, and each block stores the previous block's fingerprint — so editing any past record visibly breaks the entire chain after it.
• Consensus mechanisms make adding blocks costly (electricity in proof of work, staked capital in proof of stake), which is what prevents double-spends and cheap rewrites of history.
• Every node independently verifies every transaction, so block producers have a turn, not authority — invalid blocks are rejected regardless of who made them.
• Blockchains make tampering detectable and prohibitively expensive, not magically impossible; recent blocks firm up with confirmations.
• The ledger's integrity says nothing about the honesty of what's written on it or the safety of your keys — those risks live outside the chain.