Proposer-Builder Separation Explained: Ethereum's MEV Fix

Proposer-builder separation (PBS) is a design in Ethereum's block-production pipeline that splits the job of constructing a block's actual contents (building) from the job of proposing that block to the network (proposing), aiming to reduce the centralization pressure and technical burden that MEV extraction would otherwise place directly on validators.

Why Separating These Roles Was Necessary

Without PBS, every individual validator would, in principle, need to run sophisticated block-building infrastructure themselves in order to capture the full available MEV in each block they're assigned to propose — optimally ordering transactions, spotting arbitrage, integrating private order flow, and so on. This creates a serious problem: validators with access to better infrastructure, more capital, or more sophisticated tooling would systematically extract more value than smaller or less resourced validators, creating a strong centralizing pressure that could favor large, well-resourced staking operations over smaller and solo validators.

PBS addresses this by letting proposers outsource the complex job of block building to specialized builders, while keeping the simpler proposing role — just picking the most valuable available block and signing it — accessible to any validator, regardless of their own technical sophistication or resources.

How PBS Works in Practice

The current implementation of PBS on Ethereum largely runs through infrastructure like the one pioneered by Flashbots, operating roughly as follows:

1. Builders compete to construct full candidate blocks, incorporating both regular public transactions and privately submitted MEV-seeking bundles, aiming to maximize the total value (fees plus MEV) of the block.
2. Builders submit bids for their candidate blocks to relays, essentially saying "if you choose my block, I'll pay you (the proposer) this much."
3. Relays verify a submitted block is valid and, critically, withhold the actual detailed contents from the proposer until the proposer has already committed to it — this prevents a proposer from peeking at a lucrative block's contents and then stealing the opportunity for themselves instead of including the builder's block as submitted.
4. The proposer (validator) selects the highest-paying valid bid it's offered and signs off on that block header, without ever needing to know or evaluate the detailed transaction contents itself.

This arrangement means an individual solo validator can capture close to the same MEV-driven proposer revenue as a large, sophisticated staking operation, simply by accepting the best available bid from the competitive builder market — leveling the playing field considerably compared to a world without PBS.

Current State: Out-of-Protocol vs. In-Protocol PBS

It's worth understanding that the PBS system running today operates largely as an "out-of-protocol" arrangement — implemented through voluntary software and relay infrastructure (like MEV-Boost) that validators can choose to run, rather than being enforced directly by Ethereum's core consensus rules. Ethereum's longer-term roadmap includes moving toward "in-protocol" PBS, formalizing these roles directly at the protocol level rather than relying on optional middleware, which would reduce reliance on trusted relay operators as an intermediary. This work falls under the broader "Splurge" phase discussed in our overview of Ethereum's roadmap.

Without PBS vs. With PBS

Aspect	Without PBS	With PBS
Who builds blocks	Every validator individually	Specialized, competing builders
Barrier to capturing MEV fairly	High — favors sophisticated/well-resourced validators	Lower — any validator can accept the best bid
Centralization pressure	Favors large, well-equipped operators	Reduced, though builder-side centralization remains a risk
Current implementation	N/A	Largely out-of-protocol (e.g., via relay middleware)

Remaining Risks

PBS doesn't eliminate centralization concerns entirely — it shifts them from the proposer layer to the builder and relay layer. If a small number of builders or relays come to dominate the market, similar concentration risks re-emerge, just at a different point in the pipeline. This is an active area of ongoing research and protocol development, not a fully solved problem.

Bottom Line

Proposer-builder separation lets validators capture competitive MEV-driven revenue without needing to build sophisticated blocks themselves, by outsourcing block construction to a competitive builder market while keeping the simpler act of proposing accessible to any validator. It's a meaningful step toward reducing MEV-driven centralization pressure, though it currently relies mostly on optional, out-of-protocol infrastructure, and centralization risk hasn't disappeared so much as moved to the builder and relay layer. For the infrastructure that enabled this model, see our explainer on Flashbots.