Understanding the Bitcoin Mempool
The mempool is the waiting area for unconfirmed Bitcoin transactions. Each node maintains its own mempool, so the exact queue can vary slightly by node, but the basic job is the same: pending transactions wait until a miner selects them for a block. When transactions arrive faster than blocks can confirm them, the mempool grows and congestion follows.
Fees are expressed in satoshis per virtual byte (sats/vB). A transaction of 200 vB at 50 sats/vB costs about 10,000 satoshis. Miners rank transactions by fee relative to size and prioritize higher-fee transactions. SegWit expanded theoretical capacity, but realistic blocks still vary between 2–4 MB, leaving limited room during demand spikes.
Why Congestion Occurs
Block space is finite, so congestion appears when demand for block space outruns supply. Data-heavy uses such as Ordinals and BRC-20 minting make transactions larger, while big market moves and exchange-related transfers add sudden bursts of demand. Those dynamics push users to bid higher in fees.
Bitcoin Core’s own rules can also shape the queue. In September 2025, Bitcoin Core v29.1 reduced the default minimum relay fee from 1 sat/vB to 0.1 sat/vB. In November 2025, cluster mempool was merged for inclusion in Bitcoin Core v31.0; it changes how nodes track related transactions but does not change on-chain fees, transaction volume, or block size limits.
Mempool Conditions Through 2025
Mempool data through 2025 showed two very different phases. On Jan. 8, 2025, Johoe’s Bitcoin Mempool Statistics showed a backlog just over 200,000 transactions. By Feb. 2, 2025, a Bitcoin.com News report said mempool.space indicated 5,988 transfers in the queue. The Motley Fool reported the backlog fell from roughly 287,000 in late December 2024 to around 3,000 in early February 2025.
On Feb. 15, 2025, mempool.space showed the high-priority fee at about 1 sat/vB, or about $0.14 per transaction, and Bitcoin.com News reported block 883,885 had only 921 pending transfers. Mempool.space’s block-size report identified block 881,851 as the latest prolonged clearing event, with 3 consecutive unfilled blocks.
The quiet period extended into summer. Bitget reported the mempool had been thin since mid-May, with median fee rates down to 1 sat/vB or less and blocks often underfilled. On June 9, 2025, several blocks carried fewer than 2,000 transactions and collected only 0.01 to 0.03 BTC in total fees. CoinDesk reported that in early July the mempool was about 5,000 transactions one day and 15,000 at press time, still far below the 150,000 level seen when BTC first moved above $100,000 in late 2024.
Later in 2025, activity picked up. Bitcoin Magazine, citing CryptoQuant and FXStreet, reported the mempool expanded to about 128,000 pending transactions, its highest level since late February 2025. Blink.sv attributed the earlier near-empty mempool to low network activity and lower fees.
Monitoring the Mempool and Setting Fees
A useful monitoring routine tracks three metrics together: the median fee in sats/vB, the total size of pending transactions, and the number of pending transactions. Public tools like Mempool.space and Blockstream.info give quick fee estimates and block-level data. Running your own Bitcoin Core node provides a direct view of your node’s mempool and can serve as a fallback when browser-based tools require modern JavaScript.
When congestion is high, practical ways to control costs include consolidating UTXOs during low-fee periods, using SegWit addresses, batching payments to the same recipient, and moving small frequent payments to the Lightning Network. Replace-By-Fee (RBF) can adjust a stuck transaction, but it requires a compatible wallet and careful use.
Outlook
The 2025 pattern was episodic rather than constant: a near-empty queue at the start of the year, thin conditions through mid-year, and a rebuilding backlog toward late 2025. Because block space remains limited, fee spikes are likely whenever demand clusters around market events, ETF flows, or data-heavy applications. Watching the mempool and matching transaction timing to observed conditions remains the practical way to avoid surprise fees.
