The BIP-39 word list is a standardized set of exactly 2,048 English words from which seed phrases are constructed. Each word maps to an 11-bit number (0-2047), making a 12-word phrase encode 128 bits of entropy plus a 4-bit checksum. The list was carefully curated so that no two words share the same first four letters, reducing transcription errors.
What Is a BIP-39 Word List?
2 min read
The short version
The BIP-39 word list is a dictionary specifically designed for crypto backups. It uses common, easy-to-spell English words (like "abandon," "garden," "river") chosen so that even the first four letters uniquely identify each word. This means if you slightly mis-read your handwriting later, you can still figure out which word you meant.
How It Works
The BIP-39 standard (Bitcoin Improvement Proposal 39) defines word lists for multiple languages (English, Japanese, Spanish, Chinese, etc.). The English list contains words between 3-8 characters, chosen for: (1) uniqueness in the first 4 characters, (2) avoidance of offensive words, (3) common usage and easy spelling, (4) similarity avoidance between words (no "woman" and "women"). The encoding process: generate 128/160/192/224/256 bits of cryptographic randomness. Compute SHA-256 of the entropy and take the first entropy_bits/32 bits as checksum. Concatenate entropy + checksum. Split into 11-bit groups. Each group indexes a word. The checksum means approximately 1 in 16 random word sequences (for 12 words) would pass validation, allowing wallets to detect typos.
How 12 words encode 128 bits of entropy
A wallet generates 128 random bits: 10101001 01100110... (128 bits total). It takes SHA-256 of these bits and appends the first 4 bits as checksum, producing 132 bits. Split into twelve 11-bit segments: [10101001011, 00110100101, ...]. Each 11-bit number (0-2047) maps to a word: 1355 = "pool", 421 = "damage", etc. The resulting phrase: "pool damage artist violin rubber... " (12 words). To verify: enter the phrase in any BIP-39 wallet, it converts words back to bits, checks the checksum, and if valid, derives the master key. If you mistype one word, the checksum fails and the wallet rejects it, preventing you from accidentally using a wrong seed.
What People Get Wrong
Any 12 random words work as a seed phrase
No. The last word contains a checksum, only about 1 in 16 random word combinations are valid BIP-39 phrases. Wallets verify this checksum and reject invalid sequences.
The word order does not matter
Order is critical. "apple banana cherry" and "cherry banana apple" produce completely different private keys. Treat the sequence as a precise, ordered code.
You can substitute synonyms
Each word must be exactly as specified in the BIP-39 list. "Big" is not interchangeable with "large", only one of these is in the list, and changing it changes the underlying bits entirely.
Keep Reading
Sources & Further Reading
- BIP-39 Specification
Full technical spec covering word list generation, checksums, and seed derivation
- Ian Coleman BIP-39 Tool
Open-source tool to generate and inspect BIP-39 mnemonics (use offline only)
Questions People Also Ask
- Where can I find the full BIP-39 word list?
- The official English list is maintained in the BIP-39 GitHub repository (github.com/bitcoin/bips/blob/master/bip-0039/english.txt). It is a plain text file with 2,048 words, one per line, indexed 0-2047.
- Are BIP-39 word lists the same across all wallets?
- Yes, the English list is universal. A BIP-39 phrase generated by Ledger works in Trezor, MetaMask, or any compliant wallet. The standard exists precisely for this interoperability.
- Do other languages work the same way?
- Yes. BIP-39 defines word lists for Japanese, Korean, Spanish, Chinese (Simplified/Traditional), French, Italian, and Czech. Each follows the same encoding rules. A Japanese phrase has the same security as an English one.