Mempool — short for memory pool — is the decentralized waiting area where every unconfirmed blockchain transaction sits after being broadcast to the network, queued until a miner or validator picks it up and bundles it into the next block, with higher-fee transactions typically jumping the line. Each node on the network maintains its own version of this queue, so contents can differ slightly across peers. Think of it as a boarding queue: your ticket is valid the moment you hit send, but the highest bidder boards first.
For everyday crypto users, the mempool is the invisible force behind slow transfers and spiking network fees. When on-chain activity surges — during NFT mint rushes, DeFi liquidation cascades, or major token launches — thousands of transactions flood the queue at once. Bitcoin’s mempool depth exceeding 10 to 15 unconfirmed blocks signals severe congestion; at that point, low-fee transfers can stall for hours or even days. Reading queue conditions before you send can be the difference between a $2 fee and a $40 one.
There is a darker side to this transparency. Because pending transactions are publicly visible by default, sophisticated bots — known in DeFi circles as MEV bots — monitor the queue and insert their own orders ahead of or around yours to extract profit. This practice, called front-running or sandwich attacking, costs decentralized exchange users hundreds of millions of dollars annually. According to Ethereum.org, Miner Extractable Value (MEV) represents the total profit extractable by reordering, inserting, or censoring transactions within a block — and the pending queue is where that opportunity is first identified.
The stakes may climb higher still. In 2026, Google Quantum AI researchers published findings showing quantum circuits capable of breaking ECDLP-256 encryption using roughly 1,200 to 1,450 logical qubits — a capability that could enable mempool attacks where an adversary intercepts and redirects unconfirmed transfers within a nine-minute confirmation window. The threat is not yet practical at scale, but it underlines why public transaction visibility is both a design feature and a long-term vulnerability.
When you initiate a transfer, your wallet broadcasts the signed transaction to nearby network nodes. Each node runs basic validity checks — sufficient balance, correct digital signatures, proper formatting — and if everything passes, stores the entry in its local pending queue. Nodes then relay the data to their own peers, spreading it across the network within seconds. Because each node is autonomous and may apply different local policies, minor variations in queue contents between nodes are completely normal.
Miners and validators periodically sweep their local queue and select transactions to bundle into a new block. On Bitcoin, this is a first-price auction: miners pick the highest fee-rate entries first, measured in satoshis per virtual byte (sat/vB). As documented in the Bitcoin Wiki, transactions offering too low a fee may sit unconfirmed indefinitely during peak periods and can be evicted after a set expiry window — typically 72 hours to two weeks on default node configurations.
Tools like mempool.space provide real-time visualization of pending Bitcoin transactions, displaying fee-rate tiers, estimated confirmation times by block, and historical congestion data. Checking this dashboard before a time-sensitive send can save meaningful fees. Ethereum operates on a similar broadcast-and-hold model but uses the EIP-1559 fee structure, where a base fee is burned automatically and users optionally add a priority tip to attract validators more quickly.
A mempool attack exploits the public visibility of pending transactions to manipulate or steal value before a block is finalized. Today’s most common form is MEV-driven front-running, where bots detect large pending trades and execute their own orders first to profit from the resulting price movement. A more severe variant — identified in 2026 research by Google Quantum AI — would involve a quantum computer cracking transaction signatures within the confirmation window and redirecting funds before the block closes.
Both networks use a broadcast-and-wait model, but fee mechanics differ significantly. Bitcoin’s queue runs a straightforward fee-rate auction measured in sat/vB, while Ethereum’s EIP-1559 structure burns a base fee and lets users add optional priority tips. Ethereum’s pending pool is also more complex because smart contract interactions carry variable computational costs expressed in gas, making fee estimation less predictable than on Bitcoin’s simpler UTXO model.
Mempool.space is a free, open-source block explorer focused on Bitcoin transaction analytics. It shows pending transfers sorted by fee rate, estimated confirmation windows, current congestion heat maps, and historical fee trends. Traders and node operators use it to time submissions — sending during low-activity periods and choosing the cheapest tier that still confirms within their required window.
Understanding how the queue fits into the broader blockchain stack connects naturally to these related concepts: