Fee Market

What Is a Fee Market?

A fee market is the economic mechanism through which Bitcoin allocates its most constrained resource: block space. Every Bitcoin block can hold at most 4 million weight units of transaction data (roughly 1.5 to 2 MB of actual data depending on transaction types). When more transactions compete for inclusion than a single block can fit, an auction forms. Users attach fees to their transactions, and miners prioritize the ones paying the highest fee rate.

This market emerges naturally from Bitcoin's design. There is no central authority setting prices or managing a queue. Instead, every user independently decides how much they are willing to pay for faster confirmation, and every miner independently selects whichever transactions maximize their revenue. The result is a dynamic, decentralized price discovery mechanism for block space.

Understanding the fee market is essential for anyone transacting on Bitcoin's base layer. Overpaying wastes money; underpaying means your transaction sits in the mempool indefinitely. For a deeper analysis of fee dynamics, congestion patterns, and optimization strategies, see our Bitcoin fee market dynamics research.

How It Works

The fee market operates through three interconnected components: fee calculation, the mempool queue, and miner selection.

Fee Calculation

Bitcoin transaction fees are measured in satoshis per virtual byte (sat/vB). The formula is straightforward:

Transaction Fee = Transaction Size (vBytes) × Fee Rate (sat/vB)

Example:
  Transaction size: 140 vBytes (typical single-input SegWit transfer)
  Fee rate: 20 sat/vB
  Total fee: 140 × 20 = 2,800 satoshis (0.000028 BTC)

Virtual bytes account for the discount that SegWit introduced in 2017. Core transaction data costs 4 weight units per byte, while witness data (signatures) costs only 1 weight unit per byte. To convert weight units to virtual bytes, divide by 4. This means SegWit and Taproot transactions are cheaper than legacy formats because their signature data receives a 75% weight discount.

What matters to miners is not the total fee but the fee density: how many satoshis per virtual byte. A small transaction paying 10 sat/vB will be prioritized over a large transaction paying 5 sat/vB, because the miner earns more per unit of block space consumed.

The Mempool Queue

When a user broadcasts a transaction, it enters the mempool: a holding area of valid, unconfirmed transactions maintained by each Bitcoin node. The mempool is not a single global queue but rather a collection of local pools across the peer-to-peer network, with each node maintaining its own view.

During periods of low demand, the mempool stays small and even 1 sat/vB transactions confirm within a block or two. During congestion events, the mempool can swell to hundreds of thousands of unconfirmed transactions. In May 2023, when Ordinals inscriptions surged, the mempool ballooned to over 280,000 pending transactions totaling more than 1.5 GB of data.

Nodes enforce a default mempool size limit (typically 300 MB). When the mempool exceeds this limit, nodes evict the lowest-fee-rate transactions first, effectively setting a floor fee rate that rises with congestion.

Miner Transaction Selection

When constructing a new block template, miners run a greedy algorithm that selects transactions in descending order of fee rate until the 4-million-weight-unit limit is reached. The process is conceptually simple:

1. Sort all mempool transactions by fee rate (sat/vB), highest first
2. Add transactions to the candidate block from the top of the list
3. Account for transaction dependencies: if a transaction spends an unconfirmed output, the parent must be included first
4. Stop when the block is full or no profitable transactions remain

Modern mining software like Bitcoin Core's block assembly uses cluster mempool techniques to handle complex dependency chains more efficiently, evaluating groups of related transactions together rather than individually.

Fee Estimation

Wallets use fee estimation algorithms to predict the minimum fee rate needed for confirmation within a target number of blocks. Bitcoin Core's estimator analyzes historical confirmation data using exponentially weighted moving averages, tracking how quickly transactions at various fee rates were included in blocks.

Newer approaches like mempool-based estimators analyze the current queue depth directly, providing more responsive predictions during rapid fee changes. Most wallets offer tiered options: a high-priority rate for next-block confirmation, a medium rate for confirmation within a few blocks, and an economy rate for users willing to wait.

Fee Bumping Mechanisms

Because the fee market is dynamic, a transaction that was adequately priced at broadcast time may become underpaid if demand spikes. Bitcoin provides two mechanisms for increasing a transaction's effective fee after broadcast:

Replace-by-Fee (RBF)

Replace-by-fee allows a sender to replace an unconfirmed transaction with a new version paying a higher fee. The original transaction must signal RBF eligibility (nSequence value below 0xfffffffe per BIP 125). The replacement must pay a strictly higher absolute fee and include all original recipients at their original amounts.

# Original transaction stuck at 5 sat/vB
bitcoin-cli bumpfee <txid>

# Or create manual replacement with higher fee
bitcoin-cli createrawtransaction '[{"txid":"...","vout":0,"sequence":1}]' '{"addr":0.01}'
# Sign and broadcast the replacement at 25 sat/vB

Child-Pays-for-Parent (CPFP)

Child-pays-for-parent takes a different approach: the recipient creates a new transaction spending the stuck transaction's output, paying a high enough fee to make the combined package attractive to miners. Miners evaluate parent and child together, including both if the aggregate fee rate justifies the total weight.

CPFP is particularly useful when the recipient needs to accelerate a payment they received but cannot replace (since they did not create the original transaction).

Why It Matters

Security Budget Transition

Bitcoin's long-term security depends on the fee market. Block subsidies currently provide the majority of miner revenue: at 3.125 BTC per block after the April 2024 halving, the subsidy dwarfs typical fee revenue. As of early 2025, fees represented approximately 1 to 2% of total miner income.

But subsidies halve every 210,000 blocks (roughly four years). The next halving around April 2028 will reduce the subsidy to 1.5625 BTC. Over decades, the subsidy trends toward zero, and transaction fees must compensate. A healthy fee market is therefore not just a convenience: it is the mechanism that will sustain Bitcoin's proof-of-work security model long-term.

Layer 2 Fee Relief

Layer 2 protocols reduce fee market pressure by moving the bulk of everyday transactions off-chain. The Lightning Network enables thousands of payments to occur within payment channels, with only the opening and closing transactions touching the base layer. Similarly, Spark uses statechains to transfer Bitcoin ownership off-chain, settling on-chain only when necessary.

This creates a layered fee structure: the base layer serves as a high-value settlement network where fees reflect the security guarantee of full proof-of-work confirmation, while Layer 2 solutions handle high-frequency, lower-value payments at minimal cost. For more on how these layers compare, see our Bitcoin Layer 2 comparison research.

Historical Fee Spikes

The fee market has produced several notable congestion events that illustrate how quickly fees can escalate:

Between spikes, fees often drop to 1 to 2 sat/vB during quiet periods. This volatility underscores why fee estimation and fee bumping tools are essential for regular Bitcoin users.

Use Cases

• Transaction prioritization: users sending time-sensitive payments (exchange deposits, contract settlements) pay premium fees for next-block inclusion, while users with flexible timing can save by bidding lower
• UTXO consolidation timing: wallet operators monitor the fee market and consolidate small UTXOs during low-fee periods, reducing future transaction costs
• Transaction batching: exchanges and payment processors group multiple withdrawals into a single transaction during favorable fee conditions, amortizing the fixed overhead across many recipients
• Mining revenue optimization: miners use fee market data to decide when to broadcast their own transactions and to evaluate the profitability of fee sniping protection

Risks and Considerations

Fee Volatility

Bitcoin fees can swing from 1 sat/vB to over 500 sat/vB within hours. A transaction that costs $0.10 during a quiet weekend might cost $30 during a congestion event. This unpredictability makes Bitcoin's base layer challenging for small-value or time-sensitive payments, which is a primary motivation for Layer 2 adoption.

Security Budget Uncertainty

Whether the fee market alone can sustain sufficient hashrate security once block subsidies become negligible remains an open question. If transaction demand is insufficient or if Layer 2 solutions divert too much activity away from the base layer, the security budget could decline. This tension between scaling via Layer 2 and maintaining base-layer fee revenue is one of Bitcoin's most debated long-term design challenges.

Mempool Congestion Externalities

During extreme congestion, low-value dust UTXOs become economically unspendable because the fee required to include them in a transaction exceeds their value. This effectively destroys small balances and disproportionately affects users with many small UTXOs.

Fee Estimation Failures

Fee estimation algorithms are inherently backward-looking: they predict future fees based on recent history. Sudden demand shocks (a new protocol launch, a market crash triggering exchange withdrawals) can render estimates obsolete within minutes. Users relying on stale estimates may overpay or underpay significantly.

This glossary entry is for informational purposes only and does not constitute financial or investment advice. Always do your own research before using any protocol or technology.