Bitcoin Fee Pressure History

Pre-aggregated fee-rate stats per week bucket from our own bitcoind 28 via getblockstats. This is what fees blocks actually confirmed at — not what the mempool was predicting.

Bitcoin's fee market, briefly. Blocks 0 through roughly 350,000 (2009-2015) had effectively no fee market — most transactions paid 0 sat/vB and miners included them anyway. The first sustained mempool backlog appeared in mid-2016 after the second halving. SegWit activated in August 2017, and the first major fee spike (1,000+ sat/vB) hit in December 2017. Pressure receded through 2018-2020, returned persistently in 2021, and the inscriptions/ordinals era from early 2023 onward kept fees elevated for sustained periods. Median pre-2016 weekly buckets here sit at 0-1 sat/vB and are mostly informational; the interesting history starts around block 415,000.

Blocks in window

949,094

907 week buckets

Avg p50

46.6 sat/vB

Peak p90

3018984 sat/vB

Floor p10

0 sat/vB

State distribution (by block, across the window)

39%

26%

22%

12%

Low 39.4% (374,245 blocks) Moderate 26.4% (250,831 blocks) High 22.1% (209,954 blocks) Severe 12.0% (114,064 blocks)

Fee-rate percentiles over time

Logarithmic Y-axis — fees during ordinals storms can be 100× calm-chain rates. p10 / p50 / p90 shown by default; toggle others in the legend. Each point is one week bucket: p10 = floor (min during bucket), p50 = average median (typical block), p90 = peak (worst spike in bucket).

Window: 2008-12-29T00:00:00 to 2026-05-11T00:00:00 UTC · 949,094 blocks aggregated into 907 week buckets.

Data via the ChainQuery.com signal APIs Bitcoin Fee Pressure (GET /v1/signals/chain/block-fees/aggregate?bucket=week). Per-block detail also available at /v1/signals/chain/block-fees/history (10K-row cap).

← back to live Fee Pressure