Wine classification systems: what each one ranks

Wine classification systems rank three different things, and the front label never tells you which: a piece of ground, a company, or the months a wine sat in a barrel.

Seven systems cover most fine wine labels. The 1855 Bordeaux classification ranks châteaux, not vineyards, and has promoted one estate since 1855. Saint-Émilion ranks estates too, but reruns the exercise every decade against a published grid. Burgundy ranks parcels of land, so the rank stays with the ground when the grower changes. Champagne ranks whole villages on a percentage scale built to set grape prices. Germany's Prädikat ladder ranks sugar in the juice at harvest, which measures ripeness, not quality. Italy's DOCG, DOC and IGT rank compliance with a rulebook. Rioja's Crianza, Reserva and Gran Reserva rank ageing time. Above them all sits the EU's PDO and PGI frame. None of these systems ranks the wine in front of you, in the vintage in front of you, at the price in front of you.

What do the main wine classification systems actually rank?

Different objects, on different clocks. Read the table by column two, because that column decides whether the rank on the label can still be wrong forty years later.

System	What it ranks	The clock it runs on
1855 Bordeaux growths	The château as a name and a business	One promotion since 1855
Saint-Émilion Grand Cru Classé	The estate, scored on a published grid	Revised about every ten years; current list 2022
Burgundy grand and premier cru	The climat, a named parcel of land	33 grand cru appellations, fixed in appellation law
Champagne grand and premier cru	The village, on a percentage scale	17 grand cru villages, five of them added in 1985
German Prädikat	Must weight at harvest, in degrees Oechsle	Set by the German Wine Law of 1971
Italian DOCG, DOC and IGT	Compliance with a production rulebook	DOC from 1963, DOCG from 1980, IGT from 1992
Rioja ageing tiers	Months in oak and months in bottle	Vineyard and village categories added from 2017

Why has the 1855 Bordeaux classification barely changed?

Because it classifies names, and names do not move. Napoleon III wanted a ranking of Bordeaux for the 1855 Exposition Universelle in Paris, and the brokers who drew it up ranked the wines,'s account of the classification, according to a château's reputation and trading price.

Sixty-one red estates sit in five tiers. That same classification ranked the sweet wines of Sauternes and Barsac separately, with one Premier Cru Supérieur, eleven first growths and fifteen seconds. Château Cantemerle was added as a fifth growth in 1856. Château Mouton Rothschild moved from second to first in 1973. That is the whole edit history.

The part that matters to a buyer is structural rather than historical. This ranks châteaux, not dirt, so in that same account "the actual vineyards owned by some wineries have expanded, shrunk and been divided without any reclassification", and "as long as a vineyard parcel lies within the boundaries of the wine commune, it is eligible to be used by any chateau, independent of the quality of the parcel itself". A fifth growth can buy a neighbour's plot and sell the fruit under its classified name. A second growth can sell half its holding and keep the rank.

So the 1855 list is a snapshot of trading prices in a single year, attached to companies that have since bought and sold the land underneath it. It tells you what the brokers thought of a name before your great-grandparents were born. It does not tell you where that estate trades now, or how the wine in the bottle is showing. For the top of the list, start with the five first growths. Below that line, the label stops answering the question.

Does Saint-Émilion do it better?

It publishes its marking scheme, and the 1855 brokers never had one to publish. The Conseil des Vins de Saint-Émilion sets out the weightings in full: for a Grand Cru Classé, tasting counts for 50% of the score, reputation (advertising, distribution, value) for 20%, winery and terroirs for 20%, and winegrowing and oenology methods for 10%. A property needs 14 out of 20 to pass.

At the top tier the weighting shifts, and it shifts in an uncomfortable direction. For a Premier Grand Cru Classé, tasting stays at 50%, reputation rises to 35%, and winery and terroirs falls to 10%. A third of the score for Saint-Émilion's highest rank is awarded for being famous, which is a decent working definition of a circular test. The pass mark there is 16 out of 20.

The rest of the design beats 1855. The INAO took the classification on in 1954 and published the first list in 1955, and it has been redone roughly every decade since. The 2022 edition holds 85 estates: 2 Premiers Grands Crus Classés A, 12 Premiers Grands Crus Classés and 71 Grands Crus Classés, with Château Pavie and Château Figeac in the A category.

It is also the one estates walk out of. Château Angélus withdrew in 2021, Château Ausone and Château Cheval Blanc in 2022. And a Bordeaux court struck down the 2006 edition: it ruled on 1 July 2008 that the tasting mechanism used "was not impartial", and the 1996 list was reinstated for the 2006 to 2009 vintages. A classification you can sue, quit or fail is a classification with real information in it. It is also one whose absence proves nothing.

Does Burgundy classify the wine or the ground?

The ground, and only the ground. The BIVB lists 84 Burgundy AOCs, "more than 23% of all AOCs attributed to French wines": 7 regional appellations, 44 village appellations, premier cru climats sitting inside those villages, and 33 grand cru appellations of their own. It also corrects the folklore: "For many years, the Bourgogne has claimed to have around 100 AOCs. However, there are in fact 84 AOCs."

The tell is on the label. A grand cru carries the climat name alone, with no village attached, because the parcel outranks the commune. Corton. Montrachet. Romanée Saint-Vivant.

Which produces the opposite problem to Bordeaux's. In Bordeaux the rank follows the company and the land moves underneath it. In Burgundy the rank is welded to a few hectares that several growers divide between them, so two bottles carrying the same grand cru appellation can share a classification and nothing else: not the farming, not the élevage, not the price. The classification is real and it is not sufficient. That is why we organise the Burgundy producer atlas by grower and parcel rather than by tier.

What does a Champagne grand cru actually mean?

It means the fruit came from one of 17 communes, and it began life as a pricing formula rather than a quality judgement. The échelle des crus rated villages on a percentage scale: the lowest at 80%, premiers crus between 90 and 99%, and grands crus at 100%. The rating set the share of an agreed price a grower received for the grapes.

Seventeen villages hold grand cru status. Five were promoted in 1985: Chouilly, Le Mesnil-sur-Oger, Oger, Oiry and Verzy. Less than 9% of Champagne's planted vineyard carries the 100% rating. The pricing link has since loosened, and the ratings now inform grape negotiations rather than dictating them.

Two consequences for a buyer. First, the rank belongs to a village boundary, so a mediocre plot in Ambonnay outranks an exceptional one two hundred metres outside it. Second, most Champagne is blended across many villages and carries no cru claim at all, so the absence of "grand cru" on a Champagne label is normal rather than damning.

Why does Germany rank sugar instead of land?

Because the German Wine Law of 1971 chose a measurement it could enforce at the weighbridge. The Prädikat ladder grades must weight in degrees Oechsle: Kabinett from 67 to 82, Spätlese 76 to 90, Auslese 83 to 100, Beerenauslese and Eiswein 110 to 128, and Trockenbeerenauslese 150 to 154, with the exact thresholds varying by region and grape.

That is a ripeness scale. It is not a sweetness scale, because the winemaker decides how much of that sugar to ferment away, and it is not a site scale, because a great vineyard and an ordinary one can both deliver Spätlese must weight in a warm year.

So the growers built the missing half themselves. The VDP adopted a four-tier vineyard classification at an extraordinary general meeting in Neustadt an der Weinstraße in January 2012, effective from the 2012 vintage, and enforced it with yields: 75 hectolitres per hectare for VDP.GUTSWEIN and VDP.ORTSWEIN, 60 for VDP.ERSTE LAGE and 50 for VDP.GROSSE LAGE. Dry wines from a VDP.GROSSE LAGE site are called VDP.GROSSES GEWÄCHS and, in the VDP's words, "bottled in a special bottle embossed with the GG grape logo".

A German label can therefore carry two classifications at once: one national and legal, one private and site-based. They answer different questions and neither is redundant.

What do DOCG, DOC and IGT guarantee in Italy?

Procedure. Italy's ladder certifies that a wine obeyed a rulebook, and says almost nothing about whether the rulebook was worth obeying.

Italy launched its first official wine classification in 1963. DOCG followed in 1980. A presidential decree approved Brunello di Montalcino, Vino Nobile di Montepulciano and Barolo in July of that year, and Barbaresco in October. There are 77 DOCGs today. The step up is procedural: DOCG goes only to wines "recognized as DOC for at least seven years and are of particular quality", with chemical and sensory analysis at bottling, batch by batch, and a control band on each bottle "placed in such a way as to be damaged when the bottle is opened". DOC in turn is "exclusively reserved for wines produced in areas recognized as IGT for at least five years". IGT itself arrived in 1992.

The proof that the ladder does not rank quality is the wine at the bottom of it. Before the 1990s, Chianti could be no more than 70% Sangiovese and had to include at least 10% of a local white grape, so producers who deviated were pushed down to vino da tavola, Italy's lowest designation. Mario Incisa della Rocchetta planted Cabernet Sauvignon at Tenuta San Guido in Bolgheri in 1944, and Sassicaia was released commercially in 1971, starting with the 1968 vintage. Piero Antinori released a 1971 Sangiovese and Cabernet blend as Tignanello in 1978. Both sold outside the DOC rules, and the prices of the wines that followed them beat the prices of the best known Chianti they had been excluded from.

The law eventually followed the market: IGT Toscana in 1992, DOC Bolgheri in 1994, and Sassicaia its own Bolgheri Sassicaia DOC. The full story is in the guide to why the Super Tuscans sat outside the classification, and it is the cleanest evidence that a low rank is not a low price.

Rioja classifies time. What does that leave out?

Place. The Consejo Regulador DOCa Rioja lists its ageing tiers as a stopwatch: a red Crianza needs a minimum of two years with at least one in 225-litre oak barrels; a red Reserva needs three years, of which at least one in oak plus six months in bottle; a red Gran Reserva needs five years, two of them in oak and two in bottle. Whites and rosés run shorter at every tier.

Every one of those numbers describes what a cellar did after the harvest. None of them describes the vineyard. Two Gran Reservas can share a label category while sharing no ground, no altitude and no age of vine.

Rioja has spent the last decade adding the missing axis. The Consejo brought in Viñedo Singular, its single-vineyard category, during 2017. Vino de Zona splits the region into Rioja Alta, Rioja Alavesa and Rioja Oriental. Vino de Municipio grants naming rights to 144 municipalities. The ageing tiers stayed, because the market reads them fluently, but they are no longer the only signal on the bottle.

What sits above all of them?

The EU frame, and it is thinner than the national systems it contains. The European Commission states that for a wine PDO "the grapes have to come exclusively from the geographical area where the wine is made". For a PGI, the Commission requires that "at least 85% of the grapes used have to come exclusively from the geographical area where the wine is actually made". AOC, DOCG, DOCa and DO all sit inside that PDO category, which is why a Burgundy grand cru and a Bourgogne Aligoté sit, in Brussels terms, in the same category.

The United States declined to rank anything. Under 27 CFR 4.25, an American Viticultural Area name may appear on a label when "not less than 85 percent of the wine is derived from grapes grown within the boundaries of the viticultural area", and a state or county name needs "at least 75 percent". An AVA draws a boundary and stops there. No tiers, no growths, no revision fights.

Then there is the hole in the middle of Bordeaux. Pomerol has no official classification, yet its wines,'s summary of the appellation, "fetch very high prices at wine auctions and in the private market on a par with the most highly rated classified growths of Bordeaux". Nobody has ever ranked Pétrus, and it has never needed a rank. If a blank where the rank should sit costs nothing at auction, the rank was never the thing being priced. The market index tracks the same wines with the tier stripped out, which is the comparison that makes the point visible.

Rank the bottle, not the estate

You do not buy a classification. You buy one wine, in one vintage, at one price, and the tier on the label was awarded to a company or a commune, sometimes before your great-grandparents were born.

Every wine on this site carries a score that is 40% critic and 60% CellarTracker, so the people who opened the bottles outvote the people who tasted it once from barrel. It is the number that tells you when a fifth growth is drinking like a second, and when a famous name is coasting on a rank it was given in 1855.

The weekly note goes out with the wines whose score and whose price have moved apart. Join it and you will see the mismatches before the saleroom does.