TL;DR: A blockchain is a shared ledger built from blocks of transactions, each cryptographically linked to the block before it. Bitcoin introduced the first working blockchain in 2009 so that strangers could agree on who owns what without trusting a bank. Rewriting any past record would require redoing the proof-of-work for every later block, an energy cost that makes the history effectively permanent. Thousands of independent nodes verify every block, so no single party controls the ledger.
A Shared Ledger With No Central Keeper
Every payment system needs a ledger: a record of who paid whom, in what amount, and when. Banks, card networks, and payment apps each keep their own ledger on their own servers, and customers trust the operator to maintain it honestly. A blockchain is a way to keep that record without an operator. Thousands of computers each hold a full copy of the ledger, and a set of shared rules determines how new entries get added and why every copy stays identical. The data structure alone does not remove the operator; the surrounding design that Bitcoin built around it does.
Bitcoin shipped the first working version of this design. When Satoshi Nakamoto published the Bitcoin whitepaper in October 2008, the stated goal was a peer-to-peer electronic cash system that "would allow online payments to be sent directly from one party to another without going through a financial institution". The blockchain is the data structure that makes the goal achievable: a public, append-only history that anyone can verify and no one can quietly edit.
The word describes the structure literally. Transactions are gathered into blocks, and each block contains a hash of the previous block, a short unique fingerprint of its contents, forming a chain that runs back to the network's first block, the genesis block mined on January 3, 2009. For the broader system the ledger belongs to, see What Is Bitcoin?
What Is Inside a Block
A block has two parts: a header and a list of transactions.
The transaction list is the payload. Each transaction is a signed message that moves bitcoin from one set of addresses to another, and a typical block carries a few thousand of them. The transactions are summarized into a single hash using a Merkle tree, a structure that hashes transactions in pairs, then hashes the pairs together, until one root hash remains. Changing any transaction anywhere in the block changes the root.
The header is an 80-byte summary that contains, among other fields:
- the hash of the previous block's header, which links the block to the chain
- the Merkle root, which commits to every transaction in the block
- a timestamp and the current difficulty target
- a nonce, the value miners vary while searching for a valid proof-of-work
All of these fields are hashed with SHA-256, the same hash function used across Bitcoin for mining and address derivation. A hash function turns any input into a fixed-length output, and even a one-character change to the input produces a completely different output. That property is what makes the chain tamper-evident: any edit anywhere ripples upward into a different block hash.
How Blocks Form a Chain
Because each header contains the hash of the previous header, the blocks form an ordered sequence in which every block depends on the entire history before it. Block 900,000 commits to block 899,999, which commits to block 899,998, and so on back to the genesis block.
Suppose someone wanted to alter a transaction confirmed ten blocks ago. The edit would change that block's Merkle root, which changes its header hash, which breaks the reference stored in the next block. Repairing the break means recomputing that block too, and the one after it, all the way to the chain tip. An attacker who wants to edit one record must rebuild the entire chain from the point of the edit forward, while the network keeps extending the honest chain in the meantime.
On its own, though, hash linking only makes tampering detectable. A chain of hashes can be recomputed quickly if computing each link is cheap. The design needs a way to make each link expensive. That is what proof-of-work adds.
Proof-of-Work: The Cost of Adding a Block
To add a block to the Bitcoin blockchain, a miner must find a header whose hash falls below a network-defined difficulty target. The only known way to find one is brute force: vary the nonce, hash the header, check the result, and repeat trillions of times per second. This is proof-of-work, and the search consumes real electricity on specialized hardware.
The scheme is useful because of an asymmetry: finding a valid block hash requires enormous computation, but checking one requires a single hash operation that any laptop performs instantly. A valid header is self-certifying evidence that someone spent real resources to produce it.
The network adjusts the difficulty target every 2,016 blocks so that blocks keep arriving roughly every ten minutes regardless of how much mining hardware joins or leaves. The block interval and difficulty rules are defined in Bitcoin Core's chainparams and enforced by every node. The full mining process, including rewards and fee markets, is covered in How Does Bitcoin Mining Work?
Proof-of-work also answers a question hash linking cannot: when two valid chains exist, which one counts? Bitcoin nodes follow the chain with the most accumulated work. Honest miners build on the strongest valid chain they see, so an attacker who wants a different history must outwork the combined hashrate of every other miner on Earth, continuously, for as long as the attack lasts.
Why the Chain Resists Tampering
Rewriting a confirmed transaction requires an attacker to:
- recompute the proof-of-work for the edited block and every block after it
- do so faster than the rest of the network extends the honest chain
- produce blocks that still satisfy every consensus rule, because nodes reject invalid blocks no matter how much work they carry
Each additional block buried on top of a transaction multiplies the cost of reversing it, which is why recipients of large payments often wait for several confirmations, meaning several further blocks on top. After a handful of blocks, reversal costs more in hardware and electricity than almost any transaction is worth, and the required hashrate exceeds what any known actor controls.
The same economics explain why deeper history is, for practical purposes, settled. A transaction from 2015 sits under hundreds of thousands of blocks of accumulated work. No plausible attacker can redo that work, so the record stands without anyone needing to trust an archivist.
Full Nodes Keep the Chain Honest
Proof-of-work decides which valid chain wins, but validity itself is judged by full nodes: computers that store the blockchain and independently check every block and every transaction against the protocol rules. A node verifies that signatures are correct, that no output is spent twice, that each block's subsidy matches the issuance schedule, and that the proof-of-work meets the target.
Nodes are why the system needs no trusted operator. Miners cannot print extra bitcoin or include an invalid transaction, because every node on the network would reject the block automatically and the miner would forfeit the block reward it spent electricity to earn. Hashrate buys the right to propose blocks, and no amount of it grants authority over the rules.
Anyone can run a node on modest hardware, and tens of thousands of operators do. Independent verification is the third pillar, after hash linking and proof-of-work: the ledger is expensive to rewrite and is audited continuously by independent parties spread across jurisdictions, none of whom need permission to participate.
UTXOs: How the Bitcoin Blockchain Tracks Ownership
Bitcoin's blockchain does not store account balances. It stores transactions, and each transaction consumes previous outputs and creates new ones. An output that has not yet been spent is an unspent transaction output, or UTXO, and the set of all UTXOs, called the UTXO set, defines who can spend what at any moment. A wallet's "balance" is the sum of the UTXOs its keys control.
The UTXO model matters for the security story. Every transaction must point to specific existing outputs and carry valid signatures for them, so a node can verify any transaction against the UTXO set without trusting the sender, the miner, or any registry of accounts. Double-spending becomes a simple mechanical check: an output either exists in the set or it has already been consumed. The model is explained in depth in What Are UTXOs?
Why Bitcoin's Design Choices Make the Security Model Work
Most explanations treat "blockchain" as a free-standing technology, as if the data structure alone provided security. It does not. A chain of hashed blocks is just a tamper-evident log; what makes the Bitcoin blockchain hard to attack is the combination of design choices around it.
| Design Choice | What It Contributes | What Happens Without It |
|---|---|---|
| Proof-of-work | Makes each block costly to produce, so rewriting history carries a measurable physical price and the heaviest chain is objectively identifiable. | Block production is cheap, so an attacker can rebuild an alternative history quickly, and participants need some authority to tell them which chain is real. |
| UTXO accounting | Lets any node verify any transaction against the unspent output set, with double-spends rejected mechanically. | Validity depends on shared global account state, which is harder to verify independently and concentrates trust in whoever computes that state. |
| Cheap full nodes | Tens of thousands of independent verifiers enforce the rules, so miners hold no authority over what is valid. | Verification concentrates in a few large operators, and users must trust those operators' view of the ledger. |
| Fixed, simple rules | Conservative consensus rules and a capped supply give verifiers a stable contract to enforce decades into the future. | Frequently changing rules require users to trust whoever coordinates the changes. |
Remove any one of these and the guarantees weaken. This is why "blockchain" by itself promises little: the data structure is the easy part, and the surrounding incentive and verification design is where the security lives. Bitcoin keeps its base layer deliberately limited to preserve these properties, and pushes higher transaction volume to layers built on top, a trade-off explained in Why Does Bitcoin Use Layers?
Blockchains Beyond Bitcoin
Since 2009, thousands of other systems have adopted blockchain data structures with different rules: alternative consensus mechanisms such as proof-of-stake, flexible monetary policies, and programmable contract layers. The structure generalizes; the guarantees do not automatically follow. A blockchain secured by a small set of validators, or governed by a foundation that can change the rules, offers correspondingly weaker assurances than one secured by globally distributed proof-of-work and independent full nodes.
Private and permissioned blockchains restrict who may write to or verify the ledger. That choice reintroduces a trusted administrator, and once an administrator is trusted, a conventional replicated database usually delivers the same result with less complexity. The honest summary is that a blockchain earns its overhead when participants who do not trust each other need a shared ledger that none of them controls. Bitcoin's open monetary network fits that description; many other proposed uses do not.
What a Blockchain Cannot Do
A blockchain guarantees the integrity and ordering of the records on it. What it cannot do is vouch for the outside world: a supply-chain entry asserting that goods are authentic is only as reliable as whoever wrote it. Lost keys, mistaken payments, and user identities are also beyond its reach, and so is the trustworthiness of centralized services, since bitcoin held on an exchange is an entry in that company's internal database rather than a UTXO the customer controls on the chain.
Understanding those limits is part of understanding the design. The Bitcoin blockchain does one thing under clearly stated assumptions: it keeps a global, append-only record of bitcoin ownership that no government, company, or miner can quietly rewrite. How the full system moves value across that record is covered in How Does Bitcoin Work?
Is blockchain the same thing as Bitcoin?
No. Bitcoin is a complete monetary system, and its blockchain is the ledger that records every Bitcoin transaction. The blockchain was introduced as one component of Bitcoin's design in 2009. Other systems have since adopted the data structure, but a blockchain on its own does not provide Bitcoin's security or its fixed supply.
Who invented the blockchain?
Satoshi Nakamoto combined existing ideas, including timestamped hash chains from Haber and Stornetta and the Hashcash proof-of-work scheme, into the first working blockchain when Bitcoin launched in January 2009. The 2008 Bitcoin whitepaper describes the design, though it never actually uses the single word blockchain.
Can a blockchain be hacked or changed?
Changing a confirmed Bitcoin transaction would require redoing the proof-of-work for that block and every block after it faster than the rest of the network extends the chain. The energy cost makes this impractical, and every full node would still reject blocks that break consensus rules. Individual wallets and exchanges can be compromised, but that is a custody failure, separate from the blockchain itself.
What is stored on the Bitcoin blockchain?
Transactions. Each block contains a header and a list of transactions that move bitcoin between addresses. The blockchain does not store account balances, personal identities, or coins as files. Wallet software derives balances by scanning the set of unspent transaction outputs (UTXOs) recorded across the chain's history.
What is the difference between public and private blockchains?
A public blockchain like Bitcoin's lets anyone read the ledger, submit transactions, and run a verifying node. A private or permissioned blockchain restricts those rights to approved participants, which reintroduces a trusted administrator. With a trusted administrator in place, a conventional database usually does the same job with less overhead.