OldPubs — blog

Why the Pub Matters to the British

Mon, 28 Sep 2026 00:00:00 GMT

Britain protects its pubs in two separate ways and the two protect different things. Listing protects the fabric — section 7 of the 1990 Act is about the character of a building and says nothing about what it sells. Community-value law protects the use, for five years at a time, as a right to try rather than a right to keep. Our archive is where the two come apart: 42 closed pubs with pages of their own, several of them protected buildings with nobody behind the bar.

Of the 89 publishable pubs where our file holds both a dated building year and a first documented year of drinking, 19 show a gap of a century or more between the fabric and the trade, and nine a gap of two hundred years or more. The widest is the Tobie Norris at Stamford, where the Royal Commission reads the house as a hall and cross wing of about 1500 refronted from 1663, and the drinking begins about 2005.
We counted 42 closed pubs that have earned a page of their own in our archive, and in most of them the building has outlived the business. The Crispin and Crispianus at Strood is the plainest case: a Grade II listed shell, boarded up since a fire in March 2011 that KentOnline reported as arson, still recognisable as a pub from the street and still fully protected as a building.
The Red Lion at Wingham was registered as an asset of community value in May 2015, after a developer who had bought it from Punch applied to convert it into six flats. The application was withdrawn and the pub has not poured since. An entry on a community-value list is a right to bid when the owner sells, not an instrument that reopens a bar.
The Plough Inn at Stalisfield Green is a fifteenth-century Wealden hall house whose earliest traced record of drinking is 1841, and the Kentish Gazette of 23 October 1860 reports a coroner's inquest held on the premises. Four centuries of building stand between the frame and the first named licensee, which is why we grade the pub B and not A.

Why Do London Pubs Close at Eleven?

Mon, 28 Sep 2026 00:00:00 GMT

They do not have to, and no law says they must. Fixed national closing times ended with the Licensing Act 2003, which left each pub's hours to its own premises licence. Eleven o'clock is a habit with a paper trail, and it runs back to a 1921 statute that gave London an hour the rest of England and Wales did not get — and further back to a wartime board set up to keep munitions workers at their benches.

Eleven o'clock closing is London's before it is anybody else's, and it is datable. Section 1(1) of the Licensing Act 1921 set weekday hours at "eight hours, beginning not earlier than eleven in the morning and ending not later than ten at night, with a break of at least two hours after twelve (noon)", and then provided that "in the application of this provision to the metropolis 'nine' shall be substituted for 'eight,' and 'eleven at night' shall be substituted for 'ten at night'". The metropolis got a ninth hour in 1921; the rest of England and Wales did not get eleven until the Licensing Act 1988.
The Act that abolished closing time names no closing time. We read the consolidated text of the Licensing Act 2003 through: "permitted hours" appears five times and every one of the five is in Schedule 8, the transitional provisions, describing the hours a justices' licence allowed under the 1964 Act while old licences were being converted. The clock time "11.00 p.m." appears twice, both in Schedule 2, in the definition of late night refreshment — the supply of hot food or hot drink between 11.00 p.m. and 5.00 a.m.
Parliament abolished the wartime liquor board and kept its timetable, in the same Act and two Parts apart. Part II of the Licensing Act 1921 is headed "Winding-up of Central Control Board (Liquor Traffic)" and section 15 repeals the Intoxicating Liquor (Temporary Restriction) Act 1914 and the Defence of the Realm (Amendment) (No. 3) Act 1915 and abolishes the Board; Part I keeps the afternoon break the Board had introduced. The break was still on the statute book in section 60(1)(a) of the Licensing Act 1964 — "a break of two and a half hours beginning at three in the afternoon" — and was struck out by section 1 of the Licensing Act 1988.
Drinking-up time was doubled by an Act of Parliament whose relevant section is one sentence long. Section 2 of the Licensing Act 1988, in full: "In section 63(1) of the principal Act, for the words 'ten minutes' there shall be substituted the words 'twenty minutes'." The call that goes with it is not statutory: we read the Licensing Act 1964 end to end and the words "time, gentlemen, please" appear nowhere in it.
The "24-hour drinking" the 2003 Act was said to have brought in is, on the Home Office's own count, about one premises licence in twenty-two. At 31 March 2024 there were 224,100 premises licences in force in England and Wales and 10,300 premises with a 24-hour alcohol licence, of which 38 per cent were supermarkets and stores and 31 per cent hotel bars. The department adds that such a licence "refers to premises that are licensed for the sale or supply of alcohol 24 hours per day (but may not operate 24 hours per day)".

Why Britain Has So Many Pubs

Mon, 28 Sep 2026 00:00:00 GMT

The answer is not thirst or weather. It is licensing, decided twice: in 1552, when justices of the peace were given the power to say who could keep an alehouse, and in 1830, when Parliament took that power away again and let any ratepayer sell beer for two guineas. Within six months 24,342 people had paid. Every count since has been a subtraction from the number that produced, and our own files show the paper trail starting exactly where the licence does.

Of the 152 United Kingdom pubs in our published dataset, only five carry dated evidence from before 1552, the year justices of the peace were first required to license alehouse keepers — and not one of those five pieces of evidence is a licence. They are a taverner in Winchester's local records around 1430, a tree-ring date of 1430–32 at Norton St Philip, a scholarly dating of the fabric at Gloucester, a conveyance of 18 December 1455 at Aylesbury, and a map of Southwark of about 1542. The other forty dated records we hold for British pubs all fall in 1552 or later. We counted this on 29 September 2026 from the file below.
The claims and the evidence peak two centuries apart. Among those same 152 pubs the commonest century for a founding claim is the sixteenth, with 39 of them, while the 45 dated records we hold peak in the eighteenth, with 14. The date on a British pub sign tends to sit in the century the licensing paperwork began; the date we can document tends to sit in the century that paperwork survives from.
The Red Lion at Southampton claims 1148 for its vaulted cellar, and the earliest date Southampton's local histories give for drinking on the premises is a licence of 1552 — the same year as the Act that created licences for alehouse keepers. The claim is about a cellar and the record is about a licence, which is the shape of almost every British pub claim we hold.
Setting two pages of the Webbs' 1903 history of liquor licensing against each other gives the scale of what 1830 did: England and Wales held 51,482 licensed alehouses in 1830 (page 108), and within six months of the Beer Act taking effect on 10 October that year, 24,342 new beer-sellers had paid their two guineas to the excise (page 116). The trade acquired roughly half as many new competitors again in half a year, and the figure kept climbing — 30,978 in 1831, 33,515 in 1832, 45,717 by 1838.

The Oldest Pubs in the World

Mon, 28 Sep 2026 00:00:00 GMT

Every country with a drinking culture has a house that says it is the oldest, and the dates on the signs run from AD 560 upwards. We went through 351 of them looking for the ones where a document agrees with the sign, and the answer is not in Britain, not in Ireland, and not nearly as old as the signs would have you believe.

Of the 351 pubs with a page on this site, the oldest one whose claimed founding date is actually supported by a dated document is Taverne zur Krone at Dietikon in Switzerland. It claims 1259 and the record is 1259 — a deed of 17 October that year, read and described by Karl Grunder in the Mitteilungen der Antiquarischen Gesellschaft in Zürich for 2007. Almost no other famous claimant on this list has its claim and its record in the same century, let alone the same year.
Across the 164 pubs on this site that carry both a claimed founding year and a documented one, the median distance between the two is 28 years and the largest is 1,281. We counted this on 29 September 2026. The median is the useful number: most old pubs are roughly honest, and it is specifically the ones competing for the title that are not.
Ye Olde Fighting Cocks at St Albans claims AD 793. Historic England, writing about the question of England's oldest pub, notes that the 793 date first appears on an early-twentieth-century postcard. The earliest firm record of the house as an alehouse is 1756, under a different name, the Three Pigeons.

The Oldest Pubs in the UK

Mon, 28 Sep 2026 00:00:00 GMT

Britain makes the loudest claims about pub age and holds the youngest evidence for them. Of the 152 British pubs on this site, 107 have no dated record at all behind the year on the sign, and the oldest one that does is traced to 1430 — against a claim of 1002 made by the same house.

Of the 152 British pubs with a page on this site, only 45 carry a dated documentary record of any kind behind their founding claim. The other 107 rest on tradition, a datestone, or the house's own account. We counted this on 29 September 2026, and it is the single most useful number we hold about British pub claims.
The oldest British pub on this site with a documented record is The Royal Oak at Winchester, traced to about 1430 through records of an unnamed taverner on the site between roughly 1390 and 1430, with a 1637 schedule naming a tapster with a brewhouse. The same house claims 1002, on a legend connecting it to Queen Emma's residence.
The Cridford Inn at Trusham claims 825 and its earliest record as a pub is 1982, because that is the year it stopped being a farmhouse. CAMRA's South Devon branch and the inn's own history page both say so. The gap of 1,157 years is not a dispute about evidence: both dates are correct and they are measuring different things.

Three Capital Cities With No Page

Sat, 26 Sep 2026 00:00:00 GMT

Lisbon, Lima and Buenos Aires each have a page on this site that search engines are told to ignore. Not because the writing was poor — it was good enough to pass every check we have — but because every Portuguese pub we hold is in Lisbon, every Peruvian one in Lima and every Argentine one in Buenos Aires, so the city page and the country page are the same page wearing two addresses.

The three country regions and the three city regions on this site hold identical lists of pubs, not merely similar ones. Portugal and Lisbon both hold Cervejaria Trindade, A Ginjinha and Ginjinha Sem Rival; Peru and Lima both hold Antigua Taberna Queirolo, Bar Cordano and Piano Bar Múnich; Argentina and Buenos Aires both hold Bar El Federal, Los 36 Billares and Bar Británico. We checked the six lists against each other on 29 September 2026 and they match entry for entry.
Our own similarity check passed the three withdrawn city pages. It compares what every indexable page says once the shared furniture is stripped out, and it put the three drafts at under two per cent alike — because the prose really was different. The duplication was in what the pages were about, and a text-overlap measure cannot see that.
The three city pages were written once before, on 17 September 2026, and withdrawn the same day. They were written because one of our own counts reported them as "regions awaiting an intro" — a measurement that asked whether an indexable region was missing from the sitemap and could not tell a page held back on purpose from one nobody had got round to. The count was read as a to-do list.

The Mermaid Leaves the Top 100

Sat, 26 Sep 2026 00:00:00 GMT

The Mermaid at Rye is one of the best-known old inns in England and it has come off our own Top 100, because a topographer writing in 1847 says the house stopped being an inn about 1767 and names the private owner who had it instead. That is a break in the trade, not a gap in the paperwork, and the rule we published in September says a pub dates from when it resumed.

William Holloway's history of Rye, published in 1847, walks up Mermaid Street house by house and says of the Mermaid that "it only ceased to be an inn about eighty years ago" — about 1767 — and that its wainscoting "is carefully preserved by its present proprietor, Mr. Charles Poile". The passage names what the house was instead of an inn, which is what makes it a record of closure rather than a silence.
Boyne and Williamson's catalogue of seventeenth-century trade tokens, in its 1889 edition, says under the Rye entry for Michael Cadman's farthing that "there is now no Mermaid Inn, but it formerly stood on the north side of the present Mermaid Street at Rye, and ceased to be an inn about 1770". It is an independent statement, by different authors for a different purpose, and it agrees with Holloway to within three years.
Louis J. Jennings went to Rye to find the Mermaid in the 1870s and could not. The sexton told him "I never heerd of it"; an old man said the inn had "gone ever so long ago, sir" and pointed out the house, where "a labouring man lives in it now"; the woman at the door said "yes, but now we lives in it". Published testimony that the house was not trading, from somebody who looked.
Wikipedia's article on the Mermaid cites Holloway at page 327 for the statement that the house became a private residence. We read both scans of the 1847 edition on the Internet Archive and the passage is at page 601, in the Antiquities chapter. Anybody checking our working against that footnote will not find the sentence where the footnote sends them.
The 1891 census for Mermaid Street does not list the house as an inn because it was let as lodgings, and it appears as an inn again by the 1901 census, according to Sheila Maddock's account for the Rye Castle Museum. Between those two returns is the point at which the house starts being a pub again.

The Six Claims With a Record, Read One by One

Fri, 25 Sep 2026 00:00:00 GMT

Of the hundred oldest pub claims this site holds, 33 have a dated document of any kind behind them and six have one that reaches the year on the sign. In September this post said seven. One house has left the list because the record we held for it was its own claimed year entered twice, and that correction is the most useful thing on this page.

Of the hundred oldest claims on this site, 33 have a dated document of any kind behind them, six have a record reaching the year claimed, and 67 have nothing traced. The middle gap between claim and record is 376 years across the 33, and 409 across the 27 whose record is later than the claim.
One of the seven this post counted in September was counted in error. The New Inn at Gloucester held a first-record year of 1450, which was its own claimed year of 1450 entered a second time in the field meant for independent evidence. No document gives that year: Historic England and the Victoria County History of Gloucestershire both say mid-fifteenth-century and name none. The row is grade B and its record field is empty.
The deed behind the King's Head at Aylesbury has been traced to The National Archives' catalogue, reference 705:349/12946/494987, among the Hampton (Pakington) of Westwood Park papers at the Worcestershire Archive and Archaeology Service, which files it under King's Head Inn. Dated 18 December 1455, it is a grant from William Wandesford, esquire — not a wool merchant — to Ralph Verney, citizen and mercer of London, and five others, of three messuages called the Kyngeshede, a shop, a cellar beneath that shop and a cottage in Aylesbury.
The 1455 deeds name that house by its sign and never use the word inn. Nor does a lease of 6 November 1539, which still calls it a messuage called the Kyngs Hed. The earliest paper in the series to call it the Kinges Head Inne is a counterpart lease of 30 May 1650 to William Dawney of Aylesbury, innkeeper.
Two details this post published about Aylesbury in September are in no record we can read. That the Kyngeshede was newly built in 1455 is a National Trust guidebook's gloss repeated by Wikipedia, and the dating of the great hall to the 1470s is an English Wikipedia sentence carrying its own editors' citation-needed tag. Historic England's Grade II* list entry 1160205, which this post cited for the 1470s, mentions neither a hall nor that decade: it gives "C15-16 and 17 and later… Left hand gable C15".
Two of the six rest on charters that name an estate's tavern rather than a building — the deed of 17 October 1259 by which the Habsburgs sold Dietikon to Wettingen Abbey, which names the taverns of Dietikon and Schlieren without describing either, and the charter of 12 May 1268 in which Duke Vladislav of Opole renewed the Orlová Benedictines' rights over a taberna at Těrlicko.
One row on the hundred has a record older than its own claim: U Medvídků in Prague dates itself to 1466, and the house is recorded with a maltster from 1433 and a brewer, Jakub Starostka, from 1438.

The Photographs We Cannot Publish

Fri, 25 Sep 2026 00:00:00 GMT

Every photograph on this site prints its photographer's name and its licence, which means a photograph nobody can be credited for is a photograph we cannot use, however free it looks. Three cases we worked through this month: an archive release with the author field set to "unknown", two public-domain files released by uploaders rather than photographers, and a 1989 frame refused for a reason that had nothing to do with its licence.

The Bern City Archives prints of Gerechtigkeitsgasse 62 on Wikimedia Commons carry CC BY-SA 4.0, which requires attribution, while recording their author as "unknown" — read again on 28 September 2026. Three of the four city-archive prints offered for the Klötzlikeller are like that, so the licence asks for a credit the file cannot supply.
Two Commons photographs of the Ølhallen in Tromsø are public-domain releases made by uploaders rather than by photographers: one describes itself as "Foto: Krister Brandser" while recording the release as assumed from the uploader KEBman, and the other was uploaded by an account called Olhallen1 with the author field reading only "Own work". Both were refused here; the pub is shown instead from two CC BY 2.0 Flickr photographs.
The only photograph on Commons showing the Gleneagles Bar at Mġarr plainly — Gerd Eichmann's frame of 19 May 1989 — carries no coordinates and is described and categorised under Għajnsielem rather than Mġarr, while our record for the bar reads Mġarr, Gozo. No place-based search of ours could return it; it was found by reading the photographer's own category by hand.

The Bingley Arms: What the Directories Say

Fri, 25 Sep 2026 00:00:00 GMT

A claim to AD 953 cannot be tested against a licensing record, because England has none that early. So we went at it from the other end and read the lists that do survive for Bardsey: a trade directory of 1822, which names a publican in two other places in the same parish and none in the village, and a poll tax of 1379 that does not cover the village at all.

In Baines's directory of the county of York for 1822 the compiler named a victualler in two of the three places he listed in Bardsey parish — Richard Ellis at the Fox's Arms in Rigton and John Hudson at the Chequer Inn at Wike — and listed eleven trades at Bardsey itself without an innkeeper or victualler among them.
The poll tax of 1379 as printed for Bardsey parish covers two townships only: East Rigton, whose sum of 20s 8d is sixty-two payments of fourpence, and Wothersome, which is one entry, Thomas de Berlay esquire and his wife at 6s 8d. There is no section for the village of Bardsey, so the earliest name-by-name list of this parish does not reach the house at all.
The earliest year this site records for the Bingley Arms, 1738, is the datestone over its original blocked doorway and not a record of drink. No year for drink sold at the house has been traced by us at all, which is why the row is graded C on tradition rather than on a document.