Bitcoin Block Propagation Tracker: Network Latency Data

Block Propagation Latency Overview

Bitcoin block propagation is the time it takes for a newly mined block to reach nodes across the network. Propagation speed directly affects mining fairness, orphan rates, and the security assumptions underlying Nakamoto consensus. Slower propagation gives well-connected miners a structural advantage, since any miner who has not yet received the latest block wastes hashrate on a stale chain tip.

Propagation times have improved dramatically since Bitcoin's early years, driven by protocol-level optimizations like compact block relay (BIP 152) and dedicated relay infrastructure. The DSN Research Group at the Karlsruhe Institute of Technology has tracked these improvements: median propagation dropped from over 6 seconds in 2015 to under 1 second by 2018, while the 90th percentile fell from over 15 seconds to approximately 2 seconds over the same period.

For real-time network health metrics including node counts and block timing, see the Bitcoin network statistics dashboard.

Compact Block Relay vs Full Block Relay

The single largest improvement to block propagation came with compact block relay, specified in BIP 152 and included in Bitcoin Core since version 0.13.0 (August 2016). Instead of transmitting every transaction in a block, compact blocks send a sketch containing short 6-byte transaction identifiers. The receiving node reconstructs the full block from transactions already in its mempool, only requesting missing transactions if needed.

The bandwidth savings are substantial: a full ~1 MB block can be reconstructed from a compact block sketch of approximately 9 to 20 KB, a 97-99% reduction. More importantly, compact blocks allow relay in 0.5x round-trip time instead of 1.5x RTT at least 75% of the time, making propagation speed largely independent of block size.

Compact blocks operate in two modes. High-bandwidth mode causes selected peers to immediately relay compact blocks after verifying proof of work, prioritizing speed. Low-bandwidth mode only announces the block header hash, sending the compact block on request. Since BIP 152 is a default feature in Bitcoin Core (the dominant node implementation), adoption among reachable nodes is near-universal.

Relay Networks: History and Current State

Before compact blocks, dedicated relay networks were essential for reducing propagation delays. These networks maintain high-bandwidth connections between major mining pools and strategically placed relay nodes, bypassing the latency of Bitcoin's peer-to-peer gossip protocol.

FIBRE (Fast Internet Bitcoin Relay Engine) achieved block relay times of 100 to 300 milliseconds globally by combining UDP transport, forward error correction, and compact blocks. The original network shut down in 2017, but Localhost Research relaunched it in March 2026 with six public nodes deployed globally, rebased on Bitcoin Core v30, and an open-source monitoring dashboard tracking block reconstruction times. Falcon, developed by Cornell researchers, introduced cut-through routing but saw utilization drop to 2.65% by February 2019 and is no longer operational.

For a deeper look at how relay infrastructure and the Stratum protocol interact with mining decentralization, see our research on Stratum V2 and mining decentralization.

Geographic Distribution and Regional Latency

Block propagation is not uniform across the globe. Nodes in regions with high node density receive blocks faster, while nodes in underserved regions face structurally higher latency. Academic research from the Blockchain Layer Zero study measured average propagation delays by region.

• Europe: 3.5 seconds average (fastest, highest node density after North America)
• North America: 5.7 seconds average
• Asia: 5.7 seconds average
• Oceania: 8.5 seconds average (133% longer than Europe)
• Africa and South America: 84%+ longer than European nodes

Node distribution is heavily concentrated in the US and Europe. Bitnodes data from 2025 shows approximately 23,400 reachable Bitcoin nodes, with the United States hosting roughly 23% (~2,339 nodes) and Germany the second largest concentration (~1,367 nodes). Nodes span 181 countries, but North America and Europe together account for over 80% of reachable nodes. Over 60% of reachable nodes operate via Tor for privacy, which introduces additional latency hops.

This geographic asymmetry means that miners in Africa, South America, and Oceania face a measurable disadvantage: blocks they mine take longer to propagate, and they receive competitor blocks later, increasing their stale block rate. Relay networks like FIBRE were designed specifically to mitigate this structural imbalance.

Mempool Congestion and Reconstruction Failures

Compact block relay depends on the receiving node already having most of the block's transactions in its mempool. When mempools diverge (due to different mempool policies, congestion, or out-of-band transactions), compact block reconstruction fails and additional round-trips are required to fetch missing transactions.

This failure mode was clearly visible during two periods of heavy network congestion in late 2024. From November 8 to 15 and December 4 to 17, compact block reconstruction success rates dropped below 50%, with some periods showing only 20 to 40% success. Each reconstruction failure forces a fallback to fetching full transactions, adding hundreds of milliseconds to seconds of additional propagation delay.

The situation improved significantly after an August 2025 policy change in Bitcoin Core that reduced the minimum relay fee rate from 1 sat/vB to 0.1 sat/vB. By harmonizing what transactions nodes accept into their mempools, reconstruction success rates climbed from the 20 to 40% range back above 80% by mid-September 2025. Several proposals aim to further improve this: block template sharing (letting nodes pre-cache expected block contents), selective prefilling (including likely-missing transactions in the compact block message), and weak blocks (sharing near-miss blocks to synchronize mempools before a real block is found).

Mining Fairness and Centralization Pressure

Propagation delay creates an asymmetric advantage for large, well-connected mining pools. When a pool mines a new block, it begins working on the next block immediately while the rest of the network is still receiving and validating the previous one. This window of wasted hashrate disproportionately affects small miners and those in poorly connected regions.

Simulation studies define mining fairness as the ratio of a miner's share of stale blocks to their share of total hash power. A ratio of 1.0 is perfectly fair: a ratio above 1.0 means the miner produces more stale blocks than their hashrate would predict, indicating a propagation disadvantage. Relay networks dramatically improve fairness: simulations show stale block rates of 0.14% with relay networks versus 1.85% with standard propagation, a roughly 92% reduction.

Historical stale block data tells the story clearly. In Bitcoin's early years, 1 to 2% of all blocks were stale. By 2017 that had fallen to 60 stale blocks in the entire year. In 2018, only 5 stale blocks were recorded. By 2019, just 2. Current stale block rates are well below 0.1%, a direct result of compact blocks, improved P2P networking, and the FIBRE relay network.

This improvement matters for Bitcoin's censorship resistance. If propagation delays were still high, only large pools with private relay infrastructure could mine profitably, concentratingblock template construction in fewer hands. Faster propagation levels the playing field, though geographic disparities remain a concern.

Historical Propagation Timeline

The evolution of block propagation technology can be divided into four distinct eras, each defined by the dominant relay mechanism.

The step-by-step breakdown for a well-connected modern node: header relay takes approximately 50 ms, compact block transmission 200 to 500 ms, full validation 100 to 300 ms, and relay to the next hop another 50 ms. Total time from mining to first relay is often under 1 second for directly connected peers.

Block Size and Propagation

Before compact blocks, propagation time scaled linearly with block size, creating an incentive for miners to produce smaller (or even empty) blocks to minimize their orphan risk. Compact blocks largely eliminated this perverse incentive by decoupling propagation speed from block size.

Average block sizes have increased substantially since the introduction of ordinals and inscriptions in January 2023: from approximately 1.11 MB to 1.69 MB on average, a 52% increase. Over 99.6% of blocks at height 800,000 and above are between 3.8 and 4.0 million weight units, near the 4 MWU limit. Despite this increase, propagation times have remained stable because compact block reconstruction is independent of the raw data size: what matters is mempool synchronization, not block weight.

Layer 2 protocols built on Bitcoin benefit from fast propagation because they rely on timely block confirmations for settlement finality. Protocols like Spark and the Lightning Network depend on the base layer confirming transactions within predictable timeframes: propagation delays that extend confirmation times could affect the security assumptions of channel-based protocols.

Monitoring Block Propagation

Several public resources provide visibility into real-time and historical propagation data:

• DSN Research Group (Karlsruhe Institute of Technology): publishes long-running propagation measurements with percentile charts
• Localhost Research FIBRE dashboard: real-time block reconstruction times and relay performance metrics (launched March 2026)
• Bitcoin Optech: regularly reports on compact block reconstruction statistics and relay protocol developments
• Bitnodes: tracks global node distribution and reachability across 181 countries

Node operators can monitor their own propagation performance using Bitcoin Core's debug log, which records when each block is received, validated, and relayed. The -debug=net flag enables detailed peer-level timing information.

Frequently Asked Questions

How long does it take for a Bitcoin block to propagate?

Under normal conditions, a newly mined block reaches over 50% of the network in under 1 second and over 90% of nodes within 2 to 6 seconds. These figures reflect modern compact block relay (BIP 152), which has been the default in Bitcoin Core since 2016. In Bitcoin's early years before compact blocks, propagation to 90% of nodes took 15 to 60 seconds.

What is compact block relay and how does it speed up propagation?

Compact block relay (BIP 152) is a protocol optimization that transmits a ~9 to 20 KB sketch instead of the full ~1 to 2 MB block. The sketch contains short 6-byte transaction identifiers, and the receiving node reconstructs the full block from transactions already in its mempool. This reduces bandwidth by 97 to 99% and allows relay in 0.5x round-trip time instead of 1.5x RTT. The tradeoff is that it fails when mempools are out of sync.

Does block size affect propagation time?

With compact block relay, propagation speed is largely independent of block size. What matters is mempool synchronization: if the receiving node already has the block's transactions, a 2 MB block propagates about as fast as a 500 KB block. Before compact blocks, propagation time scaled linearly with size, which incentivized miners to produce smaller blocks.

What happens when compact block reconstruction fails?

When the receiving node is missing transactions referenced in a compact block, it must request the full transactions from the sending peer. This adds one or more additional round-trips, which can increase propagation time by hundreds of milliseconds to several seconds. During periods of heavy mempool congestion in late 2024, reconstruction success rates dropped below 50%. The August 2025 reduction of the minimum relay fee from 1 sat/vB to 0.1 sat/vB helped harmonize mempools and restored success rates above 80%.

Why does block propagation matter for mining fairness?

Miners who receive new blocks faster can begin mining the next block sooner, wasting less hashrate on stale work. Slow propagation disproportionately harms small miners and those in poorly connected regions. Before relay networks and compact blocks, stale block rates were 1 to 2%. Today they are below 0.1%, largely leveling the playing field, though geographic disparities persist.

Is the FIBRE relay network still active?

Yes. FIBRE was originally created by Matt Corallo in 2016, shut down in 2017, and relaunched by Localhost Research in March 2026. The new deployment includes six globally distributed public nodes running on Bitcoin Core v30, with an open-source monitoring dashboard. FIBRE uses UDP transport with forward error correction to achieve 100 to 300 ms block relay times globally.

How does mempool congestion affect block propagation?

Congestion causes mempools to diverge as nodes apply different eviction and fee policies. When mempools are out of sync, compact block reconstruction fails more often, requiring additional round-trips to fetch missing transactions. This was directly observed during congestion spikes in November and December 2024, when reconstruction success rates fell into the 20 to 40% range.

Which regions have the slowest block propagation?

Nodes in Africa, South America, and Oceania experience the slowest propagation, with average delays 84% to 133% longer than European nodes. This correlates with lower node density: North America and Europe host over 80% of reachable Bitcoin nodes. European nodes average 3.5 seconds, while Oceania averages 8.5 seconds.

This tool is for informational purposes only and does not constitute financial advice. Propagation data is approximate and drawn from publicly available research by the DSN Research Group, Bitcoin Optech, and academic publications. Actual propagation times vary by node configuration, network conditions, and geographic location. Always verify current data from primary monitoring sources.