The Wordle Answer Archive: How the List Really Works

Every Wordle answer is drawn from a small, human-curated list of common five-letter words, and for the first four and a half years of the game's life no answer was ever used twice. That changed in February 2026, when The New York Times began deliberately returning old answers to the rotation. Understanding how that list works, and how the public archive of past answers has changed meaning, is now a genuine part of playing the game well.

Where does the answer list come from?

Wordle runs on two separate lists. The guess list, the words the game accepts when you type them, runs to roughly thirteen thousand entries and includes plenty of obscurities. The answer list is far smaller and was hand-picked. Creator Josh Wardle and his partner, Palak Shah, started from a much larger pool and filtered it down to about 2,315 familiar, everyday words, cutting the obscure, the unpleasant and most ordinary plurals.

Since the Times bought the game in January 2022, a dedicated editor, Tracy Bennett, has overseen the sequence. The list has been lightly edited over the years: a few scheduled answers were swapped out because they were too obscure, and at least one was changed at short notice because it collided uncomfortably with the news. Answers are set in advance rather than drawn randomly each morning, which is why the puzzle can feel, occasionally, suspiciously well timed to the season. That is curation, not coincidence, and the distinction matters for how much you can predict.

Do answers repeat now?

Yes, and this is the point on which most older guides are simply out of date. From the first puzzle in June 2021 until 1 February 2026, no answer ever repeated across 1,689 puzzles. The Times then announced, in its Gameplay newsletter in late January 2026, that it would begin adding previously run words back into play. The first word to return, on 2 February 2026, was CIGAR, the answer to the very first puzzle back in June 2021. By early November 2026 roughly thirty answers had been reused, with trackers differing by one or two depending on their cut-off date. The editor has said fresh, never-used words will keep appearing alongside the repeats, so the returning words are an addition to the rotation rather than a replacement for new ones.

Where can you check past answers?

The Times itself offers an official archive of past puzzles to eligible Games subscribers, which lets you replay old days rather than merely read their answers. Beyond that, several independent sites maintain free, searchable archives of every answer with its date and puzzle number, updated daily. To give the flavour, here are the answers for the first nine days of November 2026, each cross-checked against more than one archive: VIPER (1 November, puzzle 1961), SCOOT, POUCH, NUTTY, PARCH, DRECK, PRATE, FIRST, and LOAMY (9 November, puzzle 1969). A good archive is organised by date, states its puzzle numbers, and holds back the current day's answer to avoid spoilers. If a site cannot agree with its rivals on recent answers, trust none of its older entries either.

How does the archive actually help with today's puzzle?

Used well, the archive is a probability tool, not a cheat sheet. Before the repeat era, a word found in the archive could be struck off entirely, which quietly pruned candidates in tight endgames. That logic is now weaker and more interesting. A word that appeared last week is still, realistically, not coming straight back, and a word used in the last few months remains an unlikely pick. But a word last seen in 2021 or 2022 is properly back in play, and striking it off is now a mistake: most of the confirmed repeats have been drawn from the game's earliest period.

The practical routine is short. When your solving leaves two or three candidates, check when each last appeared. Downgrade the recent ones, keep the old and the never-used in full contention, and favour the common word over the curiosity, because the list was built from everyday vocabulary. That single habit, knowing what the archive can and cannot prove, is worth more than memorising any of it. The answers of the past no longer tell you what today is not. They tell you, more softly, what today is less likely to be.