Pieter Wuille took the stage at Scaling Bitcoin in Hong Kong at 09:05 local time on December 7, 20151. That was 01:05 UTC. His talk was "Segregated witness and its impact on scalability". He made a confession, according to a transcript of the talk. “I personally dismissed this as a solution for a long time as something non-viable, until Luke-Jr discovered that it's possible to do this as a soft-fork,” Wuille said2.
Luke-Jr is the handle of Luke Dashjr, a Bitcoin Core developer. A soft fork tightens Bitcoin's rules so that older nodes still accept the new blocks3. The design already ran as a hard fork in Sidechain Elements, a sandbox for sidechain features, CoinDesk reported. This piece follows the proposal from that stage to block 481,824, and the split it left behind.
What SegWit moved
Segregated witness, or SegWit, moves signatures into a separate part of each transaction. Wuille's talk gave the reason: signatures "account for 60% of the blockchain data," yet only validating nodes need them2. Moving them out of the transaction ID also fixed malleability. Malleability let someone alter a transaction's ID before it confirmed4.
BIP141 wrote the plan down. Eric Lombrozo, Johnson Lau and Wuille are its authors, and it was assigned on December 21, 20154. Its motivation section names the Lightning Network as a beneficiary of the malleability fix.
The design counted witness bytes at a discount. “What we do is discount the witness data by 75% for block size,” Wuille told the room2. Old nodes would still see blocks of 1 megabyte or less3.
Both sides, in their own words
Did it scale Bitcoin, or split it?
“I personally dismissed this as a solution for a long time as something non-viable, until Luke-Jr discovered that it's possible to do this as a soft-fork”
“Bitcoin decided to take a road with Segregated Witness and off-chain solutions. We are trying to continue to do what Bitcoin has been doing for some time”
Three capacity numbers
Coverage on December 8, 2015, used the biggest figure. Pete Rizzo's CoinDesk report, published at 16:40 GMT, said Gregory Maxwell positioned SegWit to provide "a fourfold increase in capacity"3. The report called SegWit "a novel workaround to the block size issue".
Wuille had given a smaller number on stage. “What this normally corresponds to, with a difficult transaction load, this is around 75% capacity increase for transactions that choose to use it,” he said2. Mike Hearn gave a third figure on January 14, 2016. He called the plan “an anemic 60% capacity increase only through an accounting trick (not counting some of the bytes in each transaction)”5.
What did blocks do? mainnet.observer publishes the average block size for every day14. Averaged over each year, blocks held 949,162 bytes in 2017. That mean reached 1,188,813 bytes in 2020 and 1,702,054 bytes in 2023. The 2023 figure is 70 percent above the old 1-megabyte line.
From proposal to block 481,824
Bitcoin Core 0.13.1 shipped the soft fork on October 27, 20166. The start date for signaling was November 15, 2016. Activation needed 95 percent of blocks in one 2,016-block period, about two weeks.
Miners took nearly nine months to get there. SegWit locked in at block 479,808, mined at 12:36 UTC on August 9, 20177. Activation came at block 481,824, mined at 01:57 UTC on August 24, 20178. That first SegWit block weighed 3,954,880 weight units and held 1,866 transactions.
By then, Bitcoin had already split. Its code forked on Tuesday, August 1, 2017, creating Bitcoin Cash, Interaksyon reported the next day16. The first Bitcoin Cash coins took nearly six hours to mine. Bitcoin Magazine's interview of July 27, 2017, had asked Amaury Séchet, Bitcoin ABC's lead developer, about the new coin's plan to remove SegWit9.
Adoption by the month
mainnet.observer, an open-source refresh of transactionfee.info, counts every transaction that spends a SegWit input12. In August 2017, those transactions were 0.09 percent of the total. December 2017 closed at 10.74 percent.
Bitcoin Core's own wallet came later. Version 0.16.0, released on February 26, 2018, brought "full support for segwit in the wallet and user interfaces"10. February 2018 ran at 14.76 percent, and March 2018 at 30.33 percent12. SegWit passed half of all transactions in October 2019, at 53.65 percent.
Taproot, the next upgrade, activated at block 709,632 on November 14, 202111. In November 2021, 81.22 percent of transactions spent SegWit12. July 2023 reached 93.11 percent, and September 2026 reached 97.49 percent.
Share of Bitcoin transactions spending at least one SegWit input, per UTC calendar month, in percent
- SegWit live, Core 0.16 wallet, Taproot live
- Every other month
- Height: Share of transactions, percent
Source: mainnet.observer [12].
Read the chart as a table
| Item | Value |
|---|---|
| Aug 2017 | 0.1% |
| Sep 2017 | 3.0% |
| Oct 2017 | 10.4% |
| Nov 2017 | 10.8% |
| Dec 2017 | 10.7% |
| Jan 2018 | 11.3% |
| Feb 2018 | 14.8% |
| Mar 2018 | 30.3% |
| Apr 2018 | 33.1% |
| May 2018 | 35.6% |
| Jun 2018 | 36.3% |
| Jul 2018 | 36.8% |
| Aug 2018 | 38.5% |
| Sep 2018 | 39.8% |
| Oct 2018 | 39.6% |
| Nov 2018 | 39.3% |
| Dec 2018 | 37.3% |
| Jan 2019 | 35.9% |
| Feb 2019 | 35.8% |
| Mar 2019 | 41.1% |
| Apr 2019 | 36.1% |
| May 2019 | 38.8% |
| Jun 2019 | 40.1% |
| Jul 2019 | 39.7% |
| Aug 2019 | 38.0% |
| Sep 2019 | 48.3% |
| Oct 2019 | 53.6% |
| Nov 2019 | 51.9% |
| Dec 2019 | 54.5% |
| Jan 2020 | 55.6% |
| Feb 2020 | 55.7% |
| Mar 2020 | 52.1% |
| Apr 2020 | 48.3% |
| May 2020 | 47.1% |
| Jun 2020 | 47.7% |
| Jul 2020 | 47.3% |
| Aug 2020 | 48.4% |
| Sep 2020 | 47.1% |
| Oct 2020 | 46.4% |
| Nov 2020 | 46.9% |
| Dec 2020 | 45.9% |
| Jan 2021 | 48.4% |
| Feb 2021 | 49.6% |
| Mar 2021 | 49.3% |
| Apr 2021 | 53.0% |
| May 2021 | 54.5% |
| Jun 2021 | 66.6% |
| Jul 2021 | 75.0% |
| Aug 2021 | 77.0% |
| Sep 2021 | 78.4% |
| Oct 2021 | 79.9% |
| Nov 2021 | 81.2% |
| Dec 2021 | 82.0% |
| Jan 2022 | 82.2% |
| Feb 2022 | 82.3% |
| Mar 2022 | 82.2% |
| Apr 2022 | 82.2% |
| May 2022 | 82.8% |
| Jun 2022 | 82.9% |
| Jul 2022 | 84.1% |
| Aug 2022 | 84.0% |
| Sep 2022 | 84.0% |
| Oct 2022 | 84.1% |
| Nov 2022 | 84.2% |
| Dec 2022 | 85.1% |
| Jan 2023 | 86.3% |
| Feb 2023 | 86.7% |
| Mar 2023 | 86.7% |
| Apr 2023 | 88.9% |
| May 2023 | 94.3% |
| Jun 2023 | 92.0% |
| Jul 2023 | 93.1% |
Inputs tell the same story by type. An input is the coin a transaction spends. Legacy inputs were 97.09 percent of all inputs in 2017, against 2.88 percent for SegWit13. SegWit inputs passed legacy in 2020, at 50.71 percent to 49.27 percent.
Growth in Taproot inputs came fast. Taproot inputs went from 0.32 percent in 2022 to 22.21 percent in 202313. In 2025, legacy held 12.21 percent, SegWit version 0 held 64.08 percent and Taproot 23.69 percent.
SegWit inputs passed legacy Bitcoin inputs in 2020, 50.7 percent to 49.3 percent
Share of all Bitcoin transaction inputs by script type, per UTC calendar year, in percent
- Legacy inputs
- SegWit version 0 inputs
- Taproot inputs
- Height: Share of inputs, percent
Source: mainnet.observer [13].
Read the chart as a table
| Item | Legacy inputs | SegWit version 0 inputs | Taproot inputs |
|---|---|---|---|
| 2016 | 100.0% | 0.0% | 0.0% |
| 2017 | 97.1% | 2.9% | 0.0% |
| 2018 | 72.1% | 27.9% | 0.0% |
| 2019 | 55.1% | 44.9% | 0.0% |
| 2020 | 49.3% | 50.7% | 0.0% |
| 2021 | 37.6% | 62.4% | 0.0% |
| 2022 | 25.2% | 74.5% | 0.3% |
| 2023 | 17.3% | 60.5% | 22.2% |
| 2024 | 11.1% | 61.7% | 27.2% |
| 2025 | 12.2% | 64.1% | 23.7% |
The case for SegWit
Wuille argued that the fix reached past block size. “This directly has an effect on scalability for various network payment transaction channels and systems like lightning and others,” he said in Hong Kong2. He also listed what else it unlocked. “So segwit fixes malleability ,allows more Script upgrades, allows fraud proofs, allows pruning blocks for historical data, improves bandwidth usage for light nodes and historical sync, and it's P2SH compatible for old senders so that non-upgraders can still send funds,” he said, as transcribed.
Script upgrades, one item on that list, left a record. BIP341, which Wuille co-wrote, defines Taproot as "SegWit version 1 spending rules"15. That soft fork activated at block 709,63211. Taproot inputs held 23.69 percent of all inputs in 202513.
The case against
Jonathan Toomim, a developer and hosted mining provider, objected in Hong Kong. “My attitude is it’s ugly and awkward and that this is not a way that’s intuitive. I just see that they’re putting segregated witness here because it will be a soft fork, but that is better off as a hard fork, where it will be more elegant and safer,” he told CoinDesk3.
Amaury Séchet led the coin that left. “Bitcoin decided to take a road with Segregated Witness and off-chain solutions. We are trying to continue to do what Bitcoin has been doing for some time,” Séchet told Bitcoin Magazine9. He kept one door open. “Bitcoin Cash will probably not see SegWit in its current shape, not as a soft fork. But fixing malleability and enabling Layer 2 solutions will happen,” he said.
Séchet's interview also lists doubts about SegWit2x, a compromise plan that summer. One was "some feature of Segregated Witness hurting long-term scalability," he said9.
What to watch
Each side can point to a number. Wuille's case rests on 97.49 percent of transactions spending SegWit in September 202612. Séchet's case rests on the choice that split the chain on August 1, 2017916.
Legacy inputs still held 12.21 percent in 2025, up from 11.07 percent in 202413. Taproot's share fell from 27.22 percent to 23.69 percent over the same two years. mainnet.observer's next yearly totals will show which way each one moves.
Sources
- Scaling Bitcoin, "Scaling Bitcoin 2015 Phase II, December 6th-7th, Hong Kong" (schedule), https://scalingbitcoin.org/hongkong2015/
- Pieter Wuille, "Segregated witness and its impact on scalability" (transcript), diyhpl.us transcripts, Dec. 7, 2015, https://diyhpl.us/wiki/transcripts/scalingbitcoin/hong-kong/segregated-witness-and-its-impact-on-scalability/
- Pete Rizzo, "Is Segregated Witness the Answer to Bitcoin's Block Size Debate?," CoinDesk, Dec. 8, 2015 (Internet Archive copy), https://web.archive.org/web/20151209191346/http://www.coindesk.com/segregated-witness-bitcoin-block-size-debate/
- Eric Lombrozo, Johnson Lau and Pieter Wuille, "BIP 141: Segregated Witness (Consensus layer)," GitHub (bitcoin/bips), assigned Dec. 21, 2015, https://github.com/bitcoin/bips/blob/master/bip-0141.mediawiki
- Mike Hearn, "The resolution of the Bitcoin experiment," blog.plan99.net, Jan. 14, 2016 (Internet Archive copy), https://web.archive.org/web/20170607220742/https://blog.plan99.net/the-resolution-of-the-bitcoin-experiment-dabb30201f7
- Bitcoin Core, "Bitcoin Core 0.13.1," Oct. 27, 2016, https://bitcoincore.org/en/releases/0.13.1/
- mempool.space, "Block 479,808" (block API), read Oct. 13, 2026, https://mempool.space/api/block/0000000000000000012e6060980c6475a9a8e62a1bf44b76c5d51f707d54522c
- mempool.space, "Block 481,824" (block API), read Oct. 13, 2026, https://mempool.space/api/block/0000000000000000001c8018d9cb3b742ef25114f27563e3fc4a1902167f9893
- Aaron van Wirdum, "The Future of 'Bitcoin Cash:' An Interview with Bitcoin ABC lead developer Amaury Séchet," Bitcoin Magazine, July 27, 2017 (Internet Archive copy), https://web.archive.org/web/20170727233610/https://bitcoinmagazine.com/articles/future-bitcoin-cash-interview-bitcoin-abc-lead-developer-amaury-s%C3%A9chet/
- Bitcoin Core, "Bitcoin Core 0.16.0," Feb. 26, 2018, https://bitcoincore.org/en/releases/0.16.0/
- mempool.space, "Block 709,632" (block API), read Oct. 13, 2026, https://mempool.space/api/block/0000000000000000000687bca986194dc2c1f949318629b44bb54ec0a94d8244
- mainnet.observer, "SegWit transition," daily data read Oct. 13, 2026, https://mainnet.observer/charts/transactions-segwit-transition/
- mainnet.observer, "Input Types (by count)," daily data read Oct. 13, 2026, https://mainnet.observer/charts/inputs-types-by-count/
- mainnet.observer, "Block size in bytes," daily data read Oct. 13, 2026, https://mainnet.observer/charts/block-size/
- Pieter Wuille, Jonas Nick and Anthony Towns, "BIP 341: Taproot: SegWit version 1 spending rules," GitHub (bitcoin/bips), assigned Jan. 19, 2020, https://github.com/bitcoin/bips/blob/master/bip-0341.mediawiki
- Interaksyon, "CRYPTOCURRENCY | Bitcoin splits, but clone off to slow start," Interaksyon (Philstar), Aug. 2, 2017, https://interaksyon.philstar.com/infotek/2017/08/02/88436/cryptocurrency-bitcoin-splits-but-clone-off-to-slow-start/