6 Letter Words Ending in E

The current file contains 241 six-letter words ending in E. Of those, 104 are also in the site’s answer pool, 114 use six distinct letters, and 127 contain a repeat. The finder combines the fixed final E with exact-position, yellow-letter, gray-letter, and duplicate-count constraints.

The most useful first split is the fifth letter: -LE has 70 records, -GE 22, -TE 20, -RE 19, -SE 19, and -CE 18. These are exact spelling groups from the local JSON, not pronunciation categories or frequency estimates.

Final-E groups from the current data

The group name comes only from positions 5 and 6. ENGINE is an -NE word, because N is immediately before its final E. It is not part of the -GE group.

The spelling-shape option is not a pronunciation claim

The optional checkbox tests one rule: the final three characters must match vowel + consonant + e. In code, that is the pattern [aeiou][^aeiou]e$. It matches 126 of the 241 current records.

That rule is deliberately labeled a vowel-consonant-E spelling shape, not “silent E.” It can include words whose pronunciation does not follow a classroom long-vowel rule, and it ignores linguistic context. For example, it can match ENGINE because its final three letters are I-N-E; that says nothing about how the word is pronounced. Use the option to inspect a spelling pattern only.

Position 4 splits the list almost in half

The stored spelling-shape flag is true for 126 of the 241 records and false for the other 115. Because no record places two vowels directly before the final E, the flag has a more practical reading: position 4 holds a vowel in exactly the 126 matching records and a consonant in all 115 others. Any tile that settles whether the fourth letter is a vowel cuts the list nearly in half.

Inside the 126-record group, the last three letters concentrate into a few spelling families:

Answer-pool overlap inside each family is thin: -ATE keeps create, pirate, relate, and update; -INE keeps define, engine, and feline. E itself deserves early testing — 480 of the 682 answer-pool words contain at least one E — and the measured starter-word experiment reflects that: its entropy leader SATIRE is one of this file’s four -IRE records, alongside admire, desire, and empire.

Three shapes cover every record that fails the test

The 115 records outside the vowel-consonant-E shape are not a grab bag. They reduce to exactly three structures, and the counts add up: 65 + 34 + 16 = 115.

Consonant + LE (65 records). Most of the 70-record -LE group puts a consonant in position 4. The largest subgroups are -BLE with 18 records (babble, double, humble), -TLE with 12 (battle, gentle, little), -PLE with 10 (people, purple, simple), -DLE with 9 (candle, handle, middle), and -GLE with 6 (jungle, single, tangle). The other five -LE records — female, futile, insole, mobile, and parole — place a vowel in position 4 and count with the vowel-consonant-E group instead.

Consonant cluster + E (34 records). Two consonants precede the final E. The biggest cluster is -NGE with 10 records (change, orange, sponge), followed by -RCE with five (fierce, source), -RSE and -DGE with four each (course, sparse; bridge, grudge), -NCE and -THE with three each (bounce, chance; soothe, writhe), -RGE with two (charge, emerge), and single records for bronze, chaste, and starve. Cluster knowledge doubles as an exclusion tool. A gray N eliminates every word containing N, and that includes all 14 records with N in position 4: the ten -NGE words plus bounce, chance, stance, and bronze. The four R-led clusters — -RCE, -RSE, -RGE, and -RVE — hold 12 records between them, so a gray R clears that entire block as well.

Consonant + vowel + E (16 records). Here position 5, not position 4, is the vowel: the eight -UE records (accrue, league, pursue, rescue, tongue, torque, unique, virtue), the four -IE records (auntie, cookie, rookie, sortie), the three -EE records (coffee, decree, degree), and agorae. These 16 are the only records in the file whose fifth letter is a vowel.

Worked filter: narrow the -GE group

Suppose G is green in position 5, E is green in position 6, R was yellow in position 2, and T is gray. Enter:

• Green pattern: ____GE
• Yellow letters: R:2
• Exclude letters: T

The current 241-record file returns FORAGE, CHARGE, RAVAGE, and EMERGE. Each ends in -GE, contains R somewhere other than position 2, and contains no T. Add further board evidence before choosing among them; the coverage sort is not an answer prediction.

Worked filter: a yellow U inside the -SE group

A second scenario shows why family knowledge should guide, not decide. Suppose E is green in position 6, S is green in position 5, U was yellow in position 2, and C and L are gray. Enter:

• Green pattern: ____SE
• Yellow letters: U:2
• Exclude letters: CL

The file returns six words: AROUSE, DEFUSE, GROUSE, MISUSE, PERUSE, and UPRISE. Five of them resolve the yellow U into position 4 and belong to the nine-record -USE family; the gray C and L had already removed accuse, blouse, clause, and excuse from that family. The sixth result, UPRISE, satisfies the same constraints with U in position 1 — and it is the only one of the six in this site’s answer pool. That is the practical lesson: even when one family dominates a green pattern, a yellow letter is not committed to the family’s usual slot. Confirm position 4 directly before assuming -USE.

Counts matter when E repeats

Final E already proves at least one E. A second E tile can establish a tighter bound. If another E is green or yellow, set a minimum such as E:2. If an extra E is gray while the final E is green, a maximum of E:1 records that only one copy remains possible. A gray duplicate does not mean “exclude E” when another E tile is confirmed.

The data contains 114 six-distinct-letter records and 127 repeated-letter records. Bridge has six distinct letters; little, coffee, and breeze repeat at least one. The right branch depends on feedback, not a general rule against repeats.

Why so many English words end in E

A few spelling conventions explain most of this list, and they are worth knowing because they compress 241 words into a small set of memorable patterns.

English spelling avoids ending a word in plain V or U. That convention — not pronunciation — is why all nine -VE records (active, alcove, arrive, derive, grieve, motive, remove, starve, strive) carry a final E, and why tongue and league end in -UE. Torque and unique add the related habit of keeping Q paired with U.

A final E keeps C and G soft. In -CE and -GE words, the E signals a soft C (the s-sound in bounce) or a soft G (the j-sound in change); without it, the spelling would suggest the hard consonants of “picnic” or “slang.” This single convention accounts for the 18-record -CE group and the 22-record -GE group — 40 records from one rule.

The magic-E vowel signal is a tendency, not a law. The classroom rule — vowel, one consonant, then E usually makes the first vowel say its name — holds across much of the -ATE, -IDE, and -INE families (create, decide, define), and it is a fast way for ESL players to recall those spellings. But this list contains its own counterexamples: engine and famine keep a short I, and the unstressed endings of palate, pirate, become, and income do not rhyme with create or chrome. Trust the shape for spelling recall, not for pronunciation.

Final E is not a modern English inflection. Endings such as plural -S and past-tense -ED are added to a base word; the E in these records is instead part of the base spelling or of a larger borrowed suffix such as -IVE or -URE. That makes this list structurally different from the final-Y finder, where one derivational ending, -LY, supplies 37 of the 113 records on its own. Final E spreads across twenty fifth-letter groups; final Y concentrates. When a green tile moves from E to Y between puzzles, expect the narrowing logic to change with it.

The word-list methodology documents the licensed source, pool boundaries, score formula, and data version. For constraints that are not specific to final E, use the feedback-based candidate solver. Return to the ending-letter hub for the other published ending list.

Data review: August 23, 2026. Every count on this page was recomputed from the published ending-e.json file and the 682-word answer pool; see the source, license, and correction record.