Ethereum’s move to proof-of-stake was the headline, but the real restructuring of block production happened more quietly - through proposer-builder separation, or PBS. The idea sounds technical and narrow, but its implications reach into MEV extraction, validator economics, and the long-term question of who gets to decide what goes into a block.
What the Split Actually Does
Before PBS, validators handled the entire block production pipeline: they collected transactions from the mempool, ordered them, and proposed the resulting block. That ordering step is where MEV - maximal extractable value - lives. A validator who notices an arbitrage opportunity between two DEX prices can sequence transactions to capture it. This made sophisticated MEV extraction a significant competitive advantage for validators willing to run custom software.
PBS separates these roles. Builders - specialized entities running optimized software - assemble blocks and bid for the right to have their block proposed. Validators, now called proposers in this context, simply select the highest bid and sign off. They earn the bid value without needing to understand or implement MEV strategies themselves.
The practical implementation on Ethereum today runs through MEV-Boost, a piece of middleware that connects proposers to a relay network. Relays receive blocks from builders, verify they’re valid, and surface the highest-value option to the proposer. As of mid-2026, the vast majority of Ethereum blocks are produced through MEV-Boost rather than locally by validators.
The Centralization Problem It Created
The efficiency gains are real. Validators without MEV expertise can still earn competitive returns. But the builder market has consolidated significantly. A small number of builders account for a disproportionate share of blocks - a pattern that has been visible in on-chain data since at least 2023 and has not materially reversed.
This matters because builders, not validators, now control transaction ordering. A builder can choose to include, exclude, or reorder transactions within the constraints of protocol rules. That’s soft censorship power, even if no builder has exercised it aggressively in obvious ways.
Relays add another trust layer. The proposer doesn’t see the full block contents before committing to include it - they’re trusting the relay’s validity attestation. The relay is therefore a single point of failure and, theoretically, manipulation.
Where Enshrined PBS Fits
Ethereum researchers have proposed moving PBS directly into the protocol rather than leaving it to MEV-Boost middleware. Enshrined PBS would formalize the proposer-builder split at the consensus layer, enabling cryptographic guarantees that relays currently provide through reputation alone. Attester-proposer separation, a related concept, takes this further by distributing the attestation role.
None of this is deployed yet. But the direction is clear: Ethereum’s roadmap acknowledges that the current MEV-Boost arrangement is a temporary workaround, and the protocol needs native mechanisms to handle block construction at scale without trusting a handful of intermediaries to behave honestly.