Censorship Resistance is the ability of a blockchain to prevent any single participant or small group from permanently stopping valid transactions from being submitted, propagated, included in blocks, and confirmed by the network. It is one of the core properties that distinguishes permissionless blockchains from payment systems or databases controlled by a central operator.
In practical terms, censorship resistance means that a user who follows protocol rules and pays the required fees should have a credible path to getting a valid transaction processed, even if some validators, block producers, sequencers, or infrastructure providers refuse to cooperate.
The property is not absolute. A blockchain can be more or less censorship-resistant depending on the distribution of block production, network architecture, transaction propagation, consensus rules, and available fallback mechanisms. Temporary exclusion is also different from sustained censorship. A transaction may wait because blocks are full or its fee is too low without anyone deliberately censoring it.
For cryptocurrency networks, the important question is whether an adversary can systematically prevent a particular user, address, application, or transaction type from accessing the blockchain.
Where Blockchain Censorship Can Occur
A transaction passes through several stages before becoming part of the canonical blockchain. Censorship can occur at different points in this path.
A wallet or RPC provider might refuse to transmit the transaction. Peer-to-peer nodes might decline to propagate it. A block builder could exclude it from candidate blocks, while a validator or other block producer could refuse to propose blocks containing it.
Layer 2 networks introduce additional points of control. A centralised rollup sequencer can refuse to process a transaction even though the underlying settlement blockchain remains operational.
A simplified transaction path illustrates the problem:
- A user creates and signs a valid transaction.
- The transaction is transmitted through an RPC endpoint or directly to network peers.
- Nodes propagate it through the peer-to-peer network or other transaction channels.
- Block-production participants receive the transaction.
- A builder, validator, miner, or sequencer decides whether to include it.
- The resulting block enters the network’s consensus process.
- Additional blocks or confirmations make reversing the transaction increasingly difficult.
Blocking one stage does not necessarily produce effective censorship. If one RPC provider refuses the transaction but the user can connect to another node, the censorship attempt can be bypassed.
The problem becomes more serious when control is concentrated across critical stages or when users have no alternative route to block inclusion.
Censorship Resistance Is Not Guaranteed by Decentralisation Alone
Decentralisation and censorship resistance are closely related, but they are not identical.
A blockchain can have thousands of nodes while block production is concentrated among a relatively small number of mining pools, validators, builders, or sequencers. Most nodes may verify the chain without having meaningful influence over which transactions enter new blocks.
For censorship resistance, the distribution of actual decision-making power matters more than the raw number of network participants.
| Network Property | Main Question | Relevance to Censorship Resistance |
| Validator decentralisation | Who can propose or attest to blocks? | Reduces dependence on individual consensus participants |
| Builder decentralisation | Who constructs block contents? | Affects transaction selection and ordering |
| Network decentralisation | Can users reach multiple independent peers? | Provides alternative propagation paths |
| Sequencer decentralisation | Who orders Layer 2 transactions? | Important for normal L2 transaction inclusion |
| Client diversity | Does the network depend on one software implementation? | Reduces common infrastructure failure or policy risks |
| Geographic diversity | Where is infrastructure operated? | Reduces exposure to a single jurisdiction or outage |
| Economic decentralisation | How concentrated is stake or hash power? | Influences the cost and feasibility of coordinated censorship |
This is why censorship resistance needs to be evaluated across the entire transaction lifecycle rather than inferred from a single decentralisation metric.
Temporary Exclusion vs Systematic Censorship
Not every transaction that fails to enter the next block has been censored.
Blockchains have finite capacity. During periods of high demand, users compete for limited block space. A transaction offering a low fee may remain pending because other transactions are economically more attractive to block producers.
Transactions can also be excluded because they are invalid. An incorrect nonce, insufficient balance, invalid signature, or inadequate fee configuration can prevent execution.
Censorship involves deliberate or systematic exclusion despite the transaction satisfying the relevant protocol requirements.
The time dimension is especially important. One validator refusing a transaction for one block may have little practical effect if the next independent proposer includes it. Coordinated refusal across many consecutive block producers is much more serious.
A network can therefore tolerate some censoring participants while remaining censorship-resistant overall, provided independent participants can eventually include the affected transactions.
Economic Security Against Censorship
Blockchain consensus can make sustained censorship expensive.
In Proof of Work systems such as Bitcoin, miners determine which transactions enter the blocks they produce. A miner can refuse a transaction, but other miners can include it. Sustained censorship becomes much more difficult if the censor does not control a sufficiently large share of hash power.
Ethereum uses Proof of Stake, where validators participate in block proposal and consensus. Again, an individual participant can refuse certain transactions, but other proposers can include them.
The effectiveness of censorship therefore depends partly on how much relevant block-production power cooperates with the censor.
Economic incentives can also work against censorship. Block producers normally earn fees from included transactions. Refusing economically attractive transactions means giving up potential revenue unless the censor has another incentive that outweighs those fees.
However, economics alone cannot guarantee neutrality. Participants may censor because of legal requirements, coordinated policy, external pressure, or strategic interests. If enough economically important infrastructure providers follow the same policy, fee incentives may not be sufficient to restore inclusion.
Protocol mechanisms and participant diversity are therefore necessary complements to economic incentives.
Censorship Resistance on Layer 2
Rollups create a different censorship model because users commonly send transactions to a sequencer rather than directly to Ethereum validators.
A centralised sequencer can provide fast confirmations and efficient ordering, but it also creates an obvious point where transactions can be refused. If the sequencer is the only practical route into the Layer 2, users become dependent on its policy and availability.
Well-designed rollups can mitigate this risk through Layer 1 fallback mechanisms. Depending on the architecture, users may be able to submit transactions, messages, or withdrawal requests through contracts on the underlying settlement layer.
Such mechanisms can provide:
- a path for transactions when the normal sequencer is unavailable;
- forced inclusion after defined conditions are satisfied;
- independent withdrawal routes;
- protection against indefinite sequencer censorship;
- verifiable state and transaction data for independent recovery;
- rules limiting how long normal Layer 2 processing can be interrupted.
These fallback paths may be slower and more expensive than ordinary Layer 2 transactions. Their purpose is not performance but credible exit and censorship resistance.
A rollup with a centralised sequencer but a robust permissionless fallback can therefore have very different censorship properties from a system where users are completely dependent on one operator.
MEV, Builders and Transaction Inclusion
Modern Ethereum block production creates another censorship-resistance challenge because the validator proposing a block may not be the participant that constructed it.
Specialised builders can assemble candidate blocks and compete to have them selected by proposers. This improves block-building efficiency and gives validators access to competitive MEV revenue, but it can concentrate transaction-selection power in the builder market.
If dominant builders refuse certain transactions, proposers that consistently select the most profitable builder bids may indirectly reproduce that censorship.
This is one reason Ethereum research has explored Inclusion Lists. Such mechanisms can allow proposers to require the inclusion of eligible transactions while builders retain responsibility for optimising the rest of the block.
The goal is to prevent Proposer-Builder Separation from weakening Ethereum’s neutrality. Builders can remain specialised without receiving unrestricted control over which valid transactions are allowed to reach the chain.
This illustrates a broader principle: blockchain architecture should provide multiple independent opportunities for censorship to be bypassed.
What Can Weaken Censorship Resistance?
Censorship resistance depends on both protocol design and the real-world distribution of infrastructure.
Concentration is a recurring risk. If most users depend on the same RPC providers, most stake is controlled by a small number of operators, or most profitable blocks come from a small builder set, formal permissionlessness may not translate into practical censorship resistance.
Regulatory and geographic concentration can create similar risks. Infrastructure distributed among several companies but located under the same jurisdiction may still respond similarly to external requirements.
Private transaction channels introduce another complication. They can protect users from some forms of MEV, but they also shift order flow away from public mempools and towards specialised intermediaries. If access to valuable order flow becomes concentrated, a smaller group can gain greater influence over block construction.
Network-level attacks can also interfere with transaction propagation. If an adversary can isolate a user or block producer from honest peers, it may prevent transactions from reaching participants willing to include them.
Censorship resistance therefore cannot be reduced to one consensus rule. It emerges from the interaction between protocol incentives, networking, block production, market structure, and fallback mechanisms.
Why Censorship Resistance Matters for Cryptocurrency
A blockchain can remain technically operational while failing one of its central purposes if specific users can be permanently prevented from transacting.
This matters particularly for cryptocurrencies because permissionless access is one of their defining characteristics. Users should not need approval from a bank, payment processor, exchange, validator, or other central operator before making a protocol-valid transfer.
Censorship resistance also protects applications rather than only individual transfers. Decentralised exchanges, lending protocols, stablecoins, governance systems, bridges, and other smart contracts depend on users being able to submit transactions reliably.
The strongest practical form of censorship resistance does not require every network participant to behave neutrally. Instead, it requires enough independent participants and alternative transaction paths that refusing service at one point cannot permanently stop a valid transaction.
That is why the property remains central as blockchain architecture becomes more specialised. Validators, builders, relays, sequencers, RPC providers, and Layer 2 operators can all improve efficiency, but each can also become a point of control. Censorship-resistant design aims to ensure that none of them becomes an unavoidable gatekeeper to the blockchain.