The Golden Age

Two Weeks in April: What the 1855 Classification Actually Ranked — and Why It Never Moved

In 1855, Bordeaux's brokers ranked the Médoc by decades of prices in under two weeks for a world's fair. The list they called an unofficial sketch barely changed.

The most consequential document in the history of fine wine is a short trade letter, produced in under two weeks by men who organized no tasting to write it, and who asked that it not be taken for anything official. The liste des vins classés de la Gironde that the brokers' syndicate sent to the Bordeaux Chamber of Commerce on 18 April 1855 sorted the red wines of the Médoc — plus a single estate in the Graves — into five classes, and the sweet whites of Sauternes and Barsac into a separate hierarchy of their own. Apart from one addition that same September and one promotion in 1973, it has never changed. To understand how a rushed piece of exposition paperwork became wine's most durable ranking, start with what the brokers actually ranked: not soils, not cellars, but prices.

A world's fair and a nervous chamber of commerce

The occasion was imperial. Napoleon III's government was preparing the Exposition Universelle de Paris of 1855, France's answer to London's Crystal Palace exhibition, and a regional committee chaired by the prefect of the Gironde asked the Bordeaux Chamber of Commerce to gather samples of the department's wines for the Paris jury, with medals at stake.

The chamber's first reaction was not enthusiasm but alarm. Medals awarded on the strength of a few tasted samples could scramble a commercial order that the Bordeaux marketplace had spent generations building. Its position, as Bruno Prats of the Académie du Vin de Bordeaux has quoted it from the 1855 correspondence, was blunt: if proprietors were to use the exposition "to destroy a classification based on the experience of long years, it would be better that none of our wines figure at the exposition at all." The compromise the chamber engineered protected the hierarchy: the Gironde's wines would be exhibited collectively and kept out of the competition, communes would send samples without owners' names — and only the crus classés would appear under their own names. For that, the chamber needed the list in writing.

Thirteen days and a letter

On 5 April 1855 the chamber wrote to the Syndicat des courtiers de commerce, the sworn brokers of the Bordeaux marketplace: "We ask you to send us the very exact and very complete list of all the classed growths of the department, specifying to which of the five classes each of them belongs." The request is itself the key piece of evidence: it does not ask whether a classification should exist, or how many classes there should be. It assumes the hierarchy is already there, and asks only for it to be written down.

Thirteen days later, on 18 April, the courtiers returned their list: 57 red wines — four firsts, twelve seconds, fourteen thirds, eleven fourths, sixteen fifths — and 21 whites. They attached a disclaimer that reads, in hindsight, almost comically: "You know how delicate a matter this classification is, and how many susceptibilities it awakens; we have therefore not had the thought of drawing up an official statement of our great wines," but only, they went on, a sketch drawn from the best sources.

The list's boundaries trace the brokers' own order books. Every red wine but one came from the Médoc; the exception was Haut-Brion in the Graves, the estate that had invented the luxury-claret brand two centuries earlier and whose price history made it impossible to omit. No Saint-Émilion or Pomerol appears at all: the right bank's trade ran largely through Libourne rather than through the Bordeaux brokers whose records the list distilled. And in 1855 the now-universal label was still rare — as Prats notes, only five entries on the list (Margaux, Lafite, Latour, Beychevelle and d'Issan) carried the word "Château" at all.

A ranking of prices, not palates

Nothing was tasted. The courtiers were the marketplace's price-setting intermediaries — paid by the merchant houses to negotiate each proprietor's crop, vintage after vintage — and since the eighteenth century they had kept internal quality categories precisely in order to set prices. The mechanism, as Prats describes it, worked as a sliding scale: each year the price of the first growths was fixed according to the vintage and the state of the market, and the other classes followed at customary discounts. The 1855 classes were that price scale, written out.

Which is why the classification was old before it was born. Thomas Jefferson, touring the vineyards in 1787, already recorded four growths "of first quality" — Margaux, La Tour de Ségur, Haut-Brion and Lafite, the exact four firsts of 1855 — and noted that seconds fetched about half the firsts' price. The broker Lawton drew up his own ranking in 1815; the first published five-class classification appeared in William Franck's treatise on the wines of the Médoc in the 1820s. Nor were these lists frozen: between Franck and 1855, Ducru-Beaucaillou and Montrose climbed to second, Palmer and Calon-Ségur to third; Lawton in 1815 had even marked Haut-Brion down, wondering whether "the poor state of cultivation of its vines" explained its slipping price. The Conseil des Grands Crus Classés' own history calls the 1855 letter "the culmination of a market reality and an evolution that had been going on for more than a century." The brokers did not invent a hierarchy in thirteen days; they photographed a moving one.

Sauternes and Barsac, classified apart

The whites had their own list and their own logic. All 21 were sweet wines, and all came from Sauternes and Barsac — no dry white entered the classification at all. The hierarchy had three tiers rather than five, and a summit of one: Yquem alone was ranked premier cru supérieur, the only wine of either colour given a class to itself, above nine premiers crus and eleven deuxièmes crus. The separate treatment reflected a separate market — in the mid-nineteenth century, as Prats observes, the great liquoreux were more highly prized than most of the reds beneath the first growths.

Two changes in 170 years

The list's first correction came within months, and it proves the method. Château Cantemerle was missing from the April list for a reason the price-based system made inevitable: its owner, Caroline de Villeneuve-Durfort, had long sold her wine directly, chiefly to Dutch buyers, so its prices were invisible in the Bordeaux brokers' books. She produced her sale records, and on 16 September 1855 — the exposition still running — Cantemerle was added to the fifth growths, in manuscript, on the original list. No price record, no rank; prove the price, gain the rank.

Then nothing for 118 years. The only other change came on 21 June 1973, when Château Mouton Rothschild was promoted from second to first growth by a ministerial decision signed by Jacques Chirac, then France's minister of agriculture — the culmination of some two decades of campaigning by Baron Philippe de Rothschild, and, as the estate itself puts it, official entry into an elite to which its prices had long belonged de facto. Mouton's motto changed with its rank: Premier ne puis, second ne daigne, Mouton suis — first I cannot be, second I do not deign to be, Mouton I am — became Premier je suis, second je fus, Mouton ne change: first I am, second I was, Mouton does not change.

Never official — and never revised

Here lies the classification's central paradox. The men who wrote it insisted it was not official; no imperial decree of 1855 enacted it; the chamber had commissioned nothing more than exhibition paperwork. Prats presses the point to its logical end: strictly speaking, "the 1855 classification does not exist" — there is only a list of classed growths drawn up by the brokers at the chamber's request, whose legal existence long remained indirect, through labelling rules reserving the words cru classé for its members. Yet the absence of an official mechanism meant the absence of a revision mechanism. A ranking created by decree can be redone by decree; a ranking that merely recorded the market's verdict had no procedure for recording a new one. The sketch hardened into scripture.

Meanwhile everything it described kept moving. The classification ranks estates, and estates are property, not geology: boundaries have shifted through purchases, exchanges and divisions ever since — nowhere more than in Margaux, where holdings were already deeply interlocked in 1855. Divisions alone carried the reds from 58 growths (after Cantemerle) to 61 today, and the whites from 21 to 27, without a single new admission. Prices moved further still. By Prats's index, a second growth sold at about 88% of a first's price in 1824 and 74% in 1863; by 1970 the figure was 19%. The firsts have detached spectacularly, and spreads inside each class now dwarf the spreads between classes — a story told in detail in the economics of Bordeaux in charts.

Why, then, has it stuck? The strongest defence, argued at book length by Dewey Markham Jr., the classification's modern historian, is that the brokers' criterion was better than it looks: a century of prices is a century of accumulated judgment, thousands of repeated verdicts by buyers spending their own money — evidence no two-week tasting panel could match. The rejoinder is that 1855 froze one moment of that judgment and then protected it from all the moments that followed. Behind the argument sits the oldest debate in Bordeaux historiography: whether the classification measured terroir, or helped make it. The brokers, at least, never claimed to be measuring anything eternal. They offered a sketch drawn from the best sources — and Bordeaux has been living inside it for 170 years.