A blockchain fork is a change to the protocol rules that govern how a network validates transactions and blocks. Because a blockchain is maintained by thousands of independent participants with no central authority, any change to those rules requires coordination — and when coordination breaks down, the chain itself can split in two.
Understanding forks helps explain how cryptocurrencies evolve, why coins like Ethereum Classic exist, and how communities handle disputes that software alone cannot resolve.
Why forks happen
Every node on a blockchain network runs software that enforces a specific set of rules. These rules determine what counts as a valid block, how large blocks can be, what transactions are permitted, and how the chain resolves conflicts. As long as every node agrees on those rules, the network stays unified.
Changes become necessary for many reasons: fixing security vulnerabilities, improving performance, adding new features, or responding to a crisis. The challenge is that changing the rules requires convincing enough of the network — miners, validators, node operators, developers, and businesses — to upgrade at the same time. When they cannot agree, a fork occurs.
Soft forks: tightening the rules
A soft fork is a backwards-compatible upgrade. The new rules are a strict subset of the old rules, meaning blocks that are valid under the new rules are also valid under the old rules. Nodes that have not upgraded will still accept the upgraded blocks — they just cannot take advantage of the new features themselves.
Think of it like adding a new requirement to a contract without voiding the old signatures. Older parties see nothing unusual; newer parties simply comply with stricter terms.
How soft forks work in practice
Because soft forks only tighten constraints, a majority of miners or validators can enforce the new rules on the rest of the network even if some nodes have not upgraded. The upgraded majority will reject any block that violates the new rules, and since they represent the longer (or heavier) chain, the non-upgraded minority will eventually follow along.
Bitcoin’s SegWit upgrade in 2017 is the most studied example. It changed how transaction data is structured to increase effective capacity and fix a long-standing bug called transaction malleability. Nodes that never upgraded still saw SegWit transactions as valid — they just treated the extra data as anyone-can-spend outputs — while upgraded nodes interpreted them correctly.
Soft forks are generally considered lower-risk upgrades because they do not force a chain split. The network can transition gradually without every participant upgrading simultaneously.
Hard forks: changing the rules entirely
A hard fork is a non-backwards-compatible change. The new rules allow or require something the old rules would reject, meaning upgraded nodes and non-upgraded nodes will permanently disagree about what counts as a valid block. If both groups have significant participation, the chain splits into two independent networks sharing a common history up to the fork point.
After a hard fork, every address that held coins before the split now holds coins on both chains. The two chains then diverge permanently, each with its own development team, community, and market price.
Planned versus contentious hard forks
Not all hard forks are acrimonious. Many are scheduled upgrades where the entire community agrees to migrate together. In these cases, the old chain simply dies out because no one continues to mine or validate it.
Contentious hard forks are different. These happen when the community genuinely cannot agree — often over something fundamental like block size, monetary policy, or a specific transaction.
| Fork | Year | Chain | Cause |
|---|---|---|---|
| Ethereum / Ethereum Classic | 2016 | Ethereum | Reversal of The DAO hack funds |
| Bitcoin / Bitcoin Cash | 2017 | Bitcoin | Block size limit dispute |
| Bitcoin Cash / Bitcoin SV | 2018 | Bitcoin Cash | Further block size and protocol disagreement |
The Ethereum split is particularly instructive. After a vulnerability drained a major smart contract called The DAO, the Ethereum developers chose to hard fork the chain and effectively reverse the theft. A minority believed this violated the principle that blockchains should be immutable — that “code is law.” They continued the original chain as Ethereum Classic. Today both chains exist, governed by entirely separate teams.
What happens to your coins during a split
When a contentious hard fork produces two live chains, holders of the original asset typically receive equivalent balances on both new chains. This is because both chains inherit the same transaction history up to the fork block.
This creates a practical issue called replay attacks. In the immediate aftermath of a split, a transaction broadcast on one chain might be valid on the other as well, since the rules were identical up to that point. Developers usually address this by introducing replay protection — small rule differences that make transactions chain-specific — but not all forks implement it immediately.
If you hold coins through a fork on a self-custody wallet, you control assets on both chains from the moment of the split. If your coins are held on an exchange, the exchange decides whether to credit both assets and when. This is one practical reason many holders keep their own keys.
Miner and validator signaling
Before a fork activates, there is often a signaling period during which miners or validators announce their intent to upgrade. This gives the community visibility into whether the fork has enough support to succeed without splitting the chain.
For proof-of-work networks, miners signal by including a flag in the blocks they produce. For proof-of-stake chains, validators may signal through on-chain governance votes. If a threshold is reached within a defined window, the upgrade activates. If not, it may be delayed, modified, or abandoned.
This process is part of the broader question of consensus mechanisms — not just how blocks are validated, but how humans governing the protocol actually agree on changes.
The political dimension
Technical disagreements are rarely purely technical. Block size debates are really debates about who the network is for — small transactions for everyday users, or large settlements for institutions. Decisions about reversing hacks are really debates about the sanctity of immutability versus the pragmatic welfare of the community.
Forks are the mechanism by which irreconcilable differences resolve themselves in a system with no CEO. When persuasion fails, the network votes with hash power, stake, and ultimately with market demand. The chain with more economic activity, developer attention, and infrastructure tends to survive as the “real” version; the other persists in a diminished form or disappears entirely.
Key takeaways
- A fork is a change to the protocol rules; when nodes disagree about those rules, the chain can split.
- Soft forks tighten the rules in a backwards-compatible way and do not force a split, as long as a majority enforces them.
- Hard forks introduce rules incompatible with the old software; if both sides have support, two permanent chains emerge with the same pre-fork history.
- After a contentious hard fork, holders typically have balances on both chains, though replay protection and exchange policies complicate access.
- Famous splits — Ethereum / Ethereum Classic, Bitcoin / Bitcoin Cash — show that disagreements are as much political as technical.
- Forks are the decentralized alternative to a CEO making a final decision: they let communities diverge rather than forcing one side to capitulate.
Next up: Layer 1 vs Layer 2