A consensus algorithm is the rule a distributed network follows to agree on which new block is added to a blockchain, so every honest node ends up with the same ledger without trusting any single party. The term is often used interchangeably with consensus mechanism and consensus protocol. Ethereum’s developer docs and NIST’s Blockchain Technology Overview treat these as nested concepts rather than synonyms. Getting the distinction right separates a one-line crypto definition from a working mental model of how Bitcoin, Ethereum, and Solana stay honest.
Key Takeaways
- Consensus, according to Ethereum’s developer documentation, is an entire stack of protocols, incentives, and ideas, and requires at least 66% of nodes to agree on the global state of the network to hold.
- The consensus algorithm that still secures Bitcoin, according to Satoshi Nakamoto’s 2008 whitepaper, is essentially one-CPU-one-vote, where the majority decision is the longest chain.
- Validators, per Ethereum’s Proof-of-Stake documentation, must deposit 32 ETH into the deposit contract and run an execution client, a consensus client, and a validator client to participate.
- The Merge, according to the Ethereum Foundation, was executed on September 15, 2022, and Ethereum’s energy consumption dropped by an estimated 99.95% after the transition.
- Five consensus models are in use, per NIST Internal Report 8202, including Proof of Work, Proof of Stake, Round Robin, Proof of Authority, and Proof of Elapsed Time.
- Solana’s Anza documentation states that Proof of History is not a consensus mechanism, but it is used to improve the performance of Solana’s Proof of Stake consensus.
- Bitcoin’s network, according to the Cambridge Centre for Alternative Finance 2025 Digital Mining Industry Report, used about 0.5% of global electricity consumption annually, with about 52.4% coming from sustainable energy sources.
How Does a Consensus Algorithm Work?
A consensus algorithm is the referee for a distributed soccer match where players do not trust each other and can be swapped in at any moment. Per NIST IR 8202, the core question is determining which user publishes the next block, and the five steps below hold across Bitcoin, Ethereum, and most modern Proof-of-Stake designs.
1. A Node Broadcasts a Transaction
A user signs a transaction with their private key and sends it to a node, which gossips it to peers. Full nodes keep a mempool of pending transactions and rebroadcast until a block producer picks them up.
2. The Protocol Selects the Next Block Producer
Selection works differently across algorithms. In Proof-of-Work, any miner can try. Under Proof-of-Stake, the protocol picks a specific validator. Ethereum’s Proof-of-Stake documentation states that one validator is randomly selected to be a block proposer in every slot via RANDAO.
3. The Producer Assembles and Publishes a Candidate Block
The producer orders transactions from the mempool into a block. In Bitcoin, the miner must also find a valid hash, a process the whitepaper describes as incrementing a nonce in the block until a value is found that gives the block’s hash the required zero bits. Under Proof-of-Stake, the validator signs the block with their validator key and broadcasts it.
4. Other Nodes Validate and Attest to the Block
Every full node independently checks that the block follows the rules, every signature is valid, and no coin is double-spent. In Ethereum, a committee of validators is randomly chosen, whose votes are used to determine the validity of the block being proposed. On Bitcoin, nodes accept the block only if the proof-of-work and all transactions check out, and then start building on it.
5. Finality Is Reached
Finality is the point where a transaction is economically irreversible. Bitcoin uses probabilistic finality (after roughly six confirmations, reversal becomes vanishingly unlikely). Ethereum uses explicit finality through checkpoint voting: If a pair of checkpoints attracts votes representing at least two-thirds of the total staked ETH, the checkpoints are upgraded to finalized status.
Why Does a Consensus Algorithm Matter?
A consensus algorithm is the only thing preventing the same coin from being spent twice on a network where participants can leave, rejoin, or cheat at any moment. The NIST Blockchain Technology Overview points out that for permissionless blockchain networks, the consensus model must work even in the presence of possibly malicious users since these users might attempt to disrupt or take over the blockchain.
Security anchors every design. Bitcoin’s whitepaper makes the guarantee explicit: If a majority of CPU power is controlled by honest nodes, the honest chain will grow the fastest and outpace any competing chains. Under Proof-of-Stake, economic stake replaces hash power as the scarce resource an attacker would need to subvert the ledger.
Liveness matters just as much. Without agreement, the network splits into conflicting histories that never reconcile. Ethereum’s docs note consensus also covers potential out-of-band social coordination as a last line of defense against attacks on the network.
Cost and sustainability round out the design axis, and the two dominant algorithms diverge sharply here. The Cambridge Centre for Alternative Finance reported in 2025 that Bitcoin’s annual electricity consumption reached about 0.5% of global consumption. The same report found that about 52.4% of energy for Bitcoin mining comes from sustainable sources, which includes 9.8% nuclear and 42.6% renewables such as hydropower and wind, up from a 2022 estimate of 37.6%. Ethereum’s transition from PoW to PoS on September 15, 2022, dropped the network’s energy consumption by an estimated 99.95%.
Our crypto coverage over the last four halving cycles points to a clear pattern: consensus algorithms are increasingly picked for their regulatory and energy profile as much as for their security properties. For broader context, see the cryptocurrency coverage on CoinLaw, which tracks algorithm changes and fork events across dozens of networks.
Pros, Cons, and Risks
Every consensus algorithm trades one set of guarantees for another. No design is strictly better on every axis, which is why NIST catalogs five models rather than naming a winner.
Advantages
- Censorship resistance. Anyone with the resources can participate as a miner or validator, so no single authority can silently block transactions.
- Permissionless entry. In PoW, anyone who can run mining hardware can join. In PoS, anyone meeting the stake minimum can become a validator.
- Cryptographic auditability. Every node independently verifies every block. NIST states that each block must be valid and thus can be validated independently by each blockchain network user.
- Byzantine fault tolerance. Most public consensus algorithms tolerate a meaningful fraction of malicious or offline nodes without halting the network.
Trade-offs and Risks
- 51% attacks on PoW. Ethereum’s documentation states: To consistently create malicious yet valid blocks, a malicious miner would have needed over 51% of the network mining power to beat everyone else. Smaller PoW chains with lower hash rates are easier targets.
- Energy cost. NIST notes that puzzle-solving computations require significant resource consumption, which draws regulatory scrutiny.
- Nothing at stake in PoS. NIST flags this risk directly: a staked user could act on every such competing chain, since it is essentially free to do so. Modern PoS designs patch this with slashing penalties.
- Finality cost under attack. On Ethereum, to revert a finalized block, an attacker would commit to losing at least one-third of the total supply of staked ETH, an economic deterrent rather than a cryptographic impossibility.
Types of Consensus Algorithms
Consensus algorithms can be grouped by the scarce resource they use to pick the next block producer. NIST IR 8202 organizes them into five distinct consensus models plus a conflict resolution layer, and most real-world chains fit one of these with minor modifications.
| Algorithm | Selection method | Example network | Resource staked |
| Proof of Work (PoW) | First to solve SHA-256 puzzle wins | Bitcoin, Litecoin | Electricity + hardware |
| Proof of Stake (PoS) | Proportional to staked coins | Ethereum, Cardano | Locked cryptocurrency |
| Delegated Proof of Stake (DPoS) | Voted delegates | EOS, Tron | Coin-weighted votes |
| Proof of Authority (PoA) | Pre-approved validators | VeChain, xDai | Real-world identity |
| Proof of Elapsed Time (PoET) | Hardware-enforced random wait | Hyperledger Sawtooth | Trusted hardware |
| Round Robin | Scheduled rotation | Many permissioned chains | Implicit trust |
Source: NIST Blockchain Technology Overview, Ethereum Foundation
Proof of Work ties block publication to computational effort. NIST describes it cleanly: In the proof of work (PoW) model, a user publishes the next block by being the first to solve a computationally intensive puzzle. Bitcoin adjusts puzzle difficulty every 2016-block cycle to target a block rate of around once every ten minutes.
Proof of Stake ties publication to economic commitment. Selection is usually proportional: NIST notes that if a user had 42 % of the entire blockchain network stake, they would be chosen 42 % of the time.
DPoS, PoA, and PoET are variants for faster finality or permissioned use. DPoS lets stakers vote for a small committee of producers. PoA trusts a known set of validators and is typical in enterprise chains. PoET relies on trusted hardware enclaves to enforce a random wait time. Round Robin is the simplest model: nodes take turns on a schedule, and NIST notes that this model ensures no one node creates the majority of the blocks. Round Robin only works in permissioned settings.
Real-World Applications
The same menu of algorithms plays out differently across the major public chains. Bitcoin still runs the original Nakamoto Proof-of-Work design. Ethereum runs a Proof-of-Stake design called Gasper after its 2022 transition. Solana runs Proof-of-Stake with a timestamp layer often mistaken for consensus. The patterns we track across crypto adoption rates by country show how geography and energy policy shape which algorithms scale, while our decentralized finance ecosystem, Defi market statistics coverage, traces how PoS chains dominate smart-contract activity.
Bitcoin: Nakamoto Consensus in Practice
Bitcoin’s consensus algorithm is the one Satoshi Nakamoto laid out in the original whitepaper. Nodes race to find a SHA-256 hash below a target value, and the winner publishes the next block. The whitepaper describes the fork-resolution rule simply: The majority decision is represented by the longest chain, which has the greatest proof-of-work effort invested in it. Because the cost of redoing work scales exponentially, attacks become prohibitively expensive after a few confirmations.
Ethereum: Gasper Proof-of-Stake
Ethereum ran PoW for roughly seven years before The Merge. The Ethereum Foundation explains the upgrade involved the joining of the original execution layer of Ethereum (the Mainnet that has existed since genesis) with its new proof-of-stake consensus layer. Validators now propose and attest to blocks; miners do not. Gasper finalizes blocks through the two-thirds checkpoint voting described earlier.
Solana: Proof-of-Stake With Proof-of-History Timestamps
Solana often gets called a “Proof of History chain,” but its own Anza documentation corrects that: Proof of History is not a consensus mechanism, but it is used to improve the performance of Solana’s Proof of Stake consensus. PoH is a cryptographic timestamp that lets validators agree on transaction ordering before block consensus runs. The actual consensus is still PoS; PoH removes a synchronization bottleneck that slows other PoS chains.
Frequently Asked Questions (FAQs)
No, and the distinction matters. Ethereum’s developer docs define a consensus mechanism as the entire stack of protocols, incentives, and ideas that allow a network of nodes to agree on the state of a blockchain. A consensus algorithm (or model, in NIST’s wording) is narrower: it is the specific rule that picks the next block producer. PoW and PoS are algorithms; the full incentive, slashing, and fork-choice system around them is the mechanism.
Proof-of-Work secures Bitcoin (the dominant cryptocurrency by market capitalization). Proof-of-Stake secures Ethereum (the dominant smart-contract chain), which switched from PoW on September 15, 2022. Between them, they account for the overwhelming majority of all cryptocurrency value. Most newer chains launched after Ethereum’s transition use PoS or a PoS variant rather than PoW.
No. Without a consensus algorithm, nodes cannot agree on a single history, and the blockchain fragments into conflicting versions. Ethereum’s documentation states that at least 66% of the nodes on the network must agree on the global state of the network for consensus to hold. Permissioned chains can use simpler models like Round Robin, although some form of agreement protocol is always required.
No, despite the name. Solana’s own Anza documentation states that Proof of History is not a consensus mechanism, but it is used to improve the performance of Solana’s Proof of Stake consensus. PoH is a cryptographic timestamp that lets validators agree on transaction ordering faster; the actual agreement on which blocks are valid comes from Solana’s PoS layer.
Proof-of-Stake, Proof-of-Authority, and Round Robin all consume roughly the same energy as a modern web application, since none of them require continuous hashing. The Ethereum Foundation reports that Ethereum’s energy consumption dropped by an estimated 99.95% after The Merge. By contrast, the Cambridge Centre for Alternative Finance estimates that Bitcoin’s Proof-of-Work network consumed about 0.5% of global electricity annually in its 2025 report.
Conclusion
A consensus algorithm is the specific block-selection rule inside a larger consensus mechanism that keeps a blockchain honest, and the two terms are not interchangeable. Bitcoin’s Proof-of-Work still runs the same Nakamoto design from the original whitepaper, trading energy cost for credibly neutral security. Ethereum’s Proof-of-Stake, switched on September 15, 2022, with a 99.95% energy drop, shows that a working algorithm can be replaced on a live multi-hundred-billion-dollar network without breaking it. NIST’s five-model taxonomy remains the most durable reference for the full design space.
Based on patterns we have tracked across our crypto coverage, consensus design is shifting from a pure security debate to a regulatory and energy debate, and the algorithms that win the next cycle will be the ones that balance Byzantine fault tolerance with a footprint regulators and grid operators can accept.
