BNB Chain Activates Pasteur Hard Fork to Enhance Bridge Security

BNB Smart Chain has activated the Pasteur hard fork on its mainnet, completing a set of protocol changes designed to close weaknesses in bridge verification and validator authorization—while also aiming to pack more transactions into each block.

BNB Chain said in a Tuesday announcement that Pasteur is now live, combining three BNB Evolution Proposals (BEPs) without altering BSC’s already established 450-millisecond block time. The upgrade focuses on tighter validator handling for cross-chain operations, safer staking and governance mechanics, and a new approach to block construction during periods of network congestion.

Key takeaways

• Pasteur is live on BNB Smart Chain mainnet, tightening bridge verification and validator authorization to reduce approval and voting ambiguities.
• BEP-682 blocks duplicate validator entries during cross-chain light-block verification, improving bridge approval correctness.
• BEP-695 strengthens protections around validator key rotation, slashing, and governance voting controls.
• BEP-675 introduces a new block-building route that lets builders submit blocks after executing transactions, while validators verify and sign before final execution checks.
• BNB Chain reports an ~88% throughput increase in QANet tests, while average gas per block rose—though the team stresses these are controlled test results, not mainnet measurements.

What Pasteur changes on BSC mainnet

Pasteur brings together three protocol upgrades—BEP-682, BEP-695, and BEP-675—addressing both security and performance bottlenecks.

According to BNB Chain, BEP-682 is designed to prevent validators from being counted more than once during cross-chain light-block verification, a change intended to make bridge approval logic more robust. BEP-695 then targets the security surface around validator lifecycle operations: it updates controls tied to validator key rotation, slashing, and governance voting so that older validator authority cannot be improperly reused.

BNB Chain also said BEP-695 blocks restricted addresses from participating in voting, a targeted governance hardening that can reduce the chance of unauthorized influence during decision-making processes.

The third component, BEP-675, changes how block proposers and validators coordinate around transaction execution. Instead of forcing validators to perform repeated work inside the tight block-production window, the new route is meant to ensure execution can be handled more efficiently without sacrificing consensus verification.

Why the new block-building route matters

Under BSC’s prior block-building approach, the builder carried out transaction execution first and then submitted a proposed block to validators. Validators, before signing, would execute transactions again to confirm the block’s contents—work BNB Chain says can take time away from execution capacity inside BSC’s 450-millisecond block window.

BNB Chain argued that when blocks are hard to assemble within that short interval, blocks can end up underfilled during busy periods. Pasteur’s change is intended to reduce that waste.

BEP-675 allows builders to submit blocks they have already executed. In this flow, validators check the proposed block against consensus rules and then sign and broadcast it. BNB Chain says validators then complete full execution verification afterward—separating consensus validation from the final execution checks to better fit the timing constraints of BSC block production.

Importantly, BNB Chain said the network does not force a single method: builders can still use the older route where validators execute transactions before signing, preserving compatibility for existing operational practices while enabling the new path when it is beneficial.

Throughput gains in QANet tests, with higher block gas

BNB Chain supported the performance motivation for BEP-675 with internal testing on QANet, described by the team as a controlled environment intended to mirror BSC’s geographically distributed validator setup.

In those tests, BNB Chain reported throughput increasing by about 88%, from 1,237 to 2,324 transactions per second when using the updated block-building route. At the same time, average gas used per block rose from 46.35 million to 84.15 million. BNB Chain said the block interval and the 100-million gas limit remained unchanged.

The network team cautioned that these results were generated under controlled test conditions and were not direct mainnet measurements. Still, the pattern is directionally useful for operators and developers: the upgrade is not only about shifting workloads between builders and validators—it’s also about enabling blocks to carry more real transaction load during peak demand.

Pasteur arrives after BSC’s earlier block-time reductions

Pasteur also fits into a broader sequence of BSC upgrades aimed at reducing block times and improving operational efficiency. Earlier changes included the Maxwell hard fork, which BNB Chain says reduced average block time from 1.5 seconds to roughly 0.8 seconds in June 2025. The follow-up Fermi upgrade then brought the network down further to the current 450-millisecond cadence.

With block intervals already compressed substantially, the logic behind Pasteur’s design becomes clearer: when blocks must be produced rapidly, redundant validator-side work can become a limiting factor. Pasteur’s new builder-to-validator execution handoff is aimed at keeping consensus verification within the schedule while still performing full execution checks.

For investors and users, the practical implication is that the chain’s scaling effort is increasingly about operational fit—making the most of a fixed block time—rather than changing core time parameters again.

What to watch next after the fork

With Pasteur now live, the key items for participants are how BSC’s validator set and block-building actors adopt the new route under real network conditions, and whether the observed test gains translate into measurable improvements on mainnet during high-traffic periods. Equally important will be monitoring whether bridge verification and governance participation behave as intended with the new validator authorization and voting restrictions in place.