I have spent more than 20 years buying, cellaring and selling Bordeaux and Burgundy, and I have watched plenty of new rating systems come and go. This one is different, because it rates the people and the farming rather than the bottle. Here is how it works, every estate Michelin named in both regions, and the ones I am buying for our own cellars at United Cellars.
By Anthony Ghattas, Founder and CEO, United Cellars
Download the PDF guide A4, all 177 estates
9 Burgundy
20 Burgundy
33 Burgundy
32 Burgundy
94 Burgundy
Michelin Grapes: a rating of the producer, not the bottle
The two reference points our customers ask me about most are the 1855 Classification and critic scores. Both have a weakness I have lived with for two decades: 1855 tells you about the past, and a 95-point score tells you about one bottle in one year. Neither tells you whether the estate is run well. In 2026 a third reference arrived that tries to answer exactly that question, and if it earns the authority Michelin's restaurant stars carry, it will shape demand and pricing for years.
I will say upfront that I am a sceptic of most new rating systems. This one earned my attention because of what it measures, not who is behind it.
Estates receive one, two or three Grapes, or a "Selected" listing for producers chosen for regular review. Rather than scoring vintages, Michelin assesses the winery as a whole against five criteria: agronomy, technical mastery, identity, balance and consistency across vintages. Inspectors are Michelin employees from sommelier, critic and production backgrounds. Unlike restaurant inspections, visits are not anonymous: inspectors introduce themselves and ask to see vineyards, cellars and tasting rooms.
| Tier | Definition |
|---|---|
| Exceptional producers whose wines can be relied upon whatever the vintage. | |
| Excellent producers that stand out within their region for quality and consistency. | |
| Very good producers making wines of character, particularly in strong years. | |
| Selected | Dependable estates kept under regular review. In spirit, the Bib Gourmand of wine. |
Burgundy was unveiled in Dijon on 7 July 2026 with 94 producers. Bordeaux followed at the Cité du Vin on 7 September with 83. Michelin has said further regions will follow, and the programme sits alongside its ownership of The Wine Advocate.
Michelin Grapes vs the 1855 Classification
My view: 1855 tells you where an estate stood in the market in 1855. The Grapes tell you how an estate is farming and delivering now. They answer different questions, and I do not see one replacing the other. What the Grapes do that nothing else has is put Pomerol, Saint-Émilion and the Médoc on one contemporary scale, which is how I have always had to think about Bordeaux as a buyer anyway.
| 1855 Classification | Michelin Grapes | |
|---|---|---|
| Basis | Reputation and market price at the time of the Paris Universal Exhibition | Five criteria: farming, winemaking, identity, balance, consistency |
| What is rated | The château name and its historic price tier | The estate as it operates today |
| Scope | Médoc reds, Haut-Brion, Sauternes and Barsac | Adds Saint-Émilion and Pomerol, so both banks sit on one ladder |
| Revisions | One change in 170 years (Mouton, 1973) | Reviewed annually |
| Viticulture | Not considered | A core criterion; Michelin credits cover crops, plot-by-plot work and organic and biodynamic conversion |
| Legal status | Fixed and protected | Editorial; changes no château's classified status |
Bordeaux
Michelin Grapes 2026 · 83 estates
83 estates across the Médoc communes, Pessac-Léognan, Saint-Émilion, Pomerol, Sauternes and Barsac. The chip on each card shows the estate's 1855 rank where one exists; Right Bank estates were never part of 1855. Gold tags are my cellar picks, charcoal tags are estates already in my own cellar, and "Shop" links go to estates we currently have in stock at United Cellars.
Three Grapes · 9
Two Grapes · 16
One Grape · 37
Selected · 21
Burgundy
Michelin Grapes 2026 · 94 estates
94 estates across the Côte de Nuits, Côte de Beaune and Côte Chalonnaise; Chablis and the Mâconnais are not covered in this first edition. Burgundy has no 1855: its hierarchy is the vineyard, not the producer, which is exactly why a producer rating is useful here. Each card shows the highest vineyard tier the estate owns or farms. GC = Grand Cru, 1er = Premier Cru. Gold tags are my picks; charcoal tags are already in my own cellar.
Three Grapes · 9
Two Grapes · 20
One Grape · 33
Selected · 32
What stands out to me in Burgundy
- I put the holdings on every card for a reason. They tell the story. Of the nine three-Grape estates, Bouley owns no Grand Cru and Lamy only a sliver of Criots-Bâtard. Michelin is rating farming and consistency, not vineyard pedigree.
- Armand Rousseau, among the most collectible domaines in Burgundy, landed at one Grape, alongside de Vogüé, Cathiard, Méo-Camuzet and Roulot. Ramonet is Selected only. All own Grand Cru land trading at four figures a bottle.
- Domaine de la Romanée-Conti is where it should be. It is the most famous name in Burgundy, arguably in all of wine, and the one estate whose absence from the top tier would have sunk the whole exercise. That it sits there on merit, farmed biodynamically across 28 hectares of Grand Cru, is the anchor the rest of the list hangs from.
- Hubert Lamy of Saint-Aubin sits with DRC, Leroy and Coche-Dury, the clearest sign that price is not being used as a proxy for quality.
- Dujac at two Grapes is the one I would argue with. Jacques Seysses built one of the most consistent domaines in the Côte de Nuits, and I have bought it for years. It goes in as a pick regardless.
- Mercurey and Rully each place a producer at two Grapes, ahead of dozens of Côte d'Or Grand Cru owners.
- Michelin's international director said the selection shows excellence is not defined solely by the prestige of a name.
- No Chablis, no Mâconnais. Michelin drew its Burgundy line at the Côte de Nuits, Côte de Beaune and Côte Chalonnaise. That leaves out Raveneau, Dauvissat and a generation of serious growers in Chablis, and the Mâconnais estates, Guffens-Heynen and the Bret Brothers among them, that now make some of the best-value white Burgundy there is. Michelin has not said why. My guess is a first edition kept to the Côte d'Or where its inspectors had done the visits, and the omission is a practical one rather than a judgement. Either way, until those regions are covered the Burgundy list is a Côte d'Or list, and it should be read as one.
- The process drew pushback: Arnoux-Lachaux received one Grape and asked to be removed, saying it had not received Michelin or presented wines to the press since 2020. A real question mark over how some estates were assessed.
Burgundy: what I am putting in the cellar
- Hubert Lamy. Saint-Aubin 1er Cru at a fraction of Puligny prices. Haute Densité is already cult; buy the standard 1er Crus (En Remilly, Derrière chez Edouard) in volume.
- Jean-Marc & Thomas Bouley. Volnay and Pommard 1er Cru only, now rated alongside DRC. The most affordable three-Grape red in either region.
- Bruno Lorenzon and Dureuil-Janthial. Côte Chalonnaise at two Grapes. The best value red and white in the entire selection.
- Paul Pillot, Lamy-Caillat, Benoît Ente, Benoît Moreau, Jean-Marc Vincent, Domaine des Croix, Gérard Mugneret, J-C Bachelet. Two-Grape estates rated above famous Grand Cru owners, still priced below them.
- Duroché, Denis Bachelet, Henri Germain, Bernard-Bonin, Roblet-Monnot, Marc Colin, Château de la Tour. One-Grape producers whose prices have not caught up.
- Selected tier: Sylvain Pataille (Marsannay before the Premiers Crus arrive), Berthaut-Gerbet, Pierre Girardin, Yvon Clerget, Bachelet-Monnot, Joseph Colin, Camille Thiriet, Maxime Cottenceau. The next generation, and where 2027's Grapes come from.
- The contrarian angle: Rousseau, de Vogüé, Cathiard, Méo-Camuzet, Roulot, Ramonet, Fourrier, Hubert Lignier. A low tier changes nothing about their Grand Cru holdings. If newer buyers hesitate, established collectors take the allocation.
Michelin Grapes as a second lens, not a replacement
I have been asked by customers, suppliers and a few of our own team whether this changes how United Cellars buys. The short answer is yes, at the margins, and mostly in the tiers below three Grapes. Here is how I am thinking about it.
- The Grapes are a producer-consistency signal. They are not a substitute for vintage scores and will not tell you which bottle to open tonight.
- The most useful information is in the one- and two-Grape tiers and the Selected list, where strong estates without first-growth or DRC price tags are identified. That is where value sits for our customers, and where we are directing our buying.
- Treat omissions cautiously. Michelin has said absence reflects an unfinished evaluation, not a poor one. Expect those names in the 2027 edition.
- 1855 is not going anywhere. It still drives en primeur pricing and the négociant system. The Grapes are a second lens, one that rewards estates on an upward trajectory and penalises complacency.
- Give it three or four editions. Michelin's restaurant stars took decades to earn their authority. The Grapes will need the same, but the direction of travel is clear: farming, precision and consistency are being rewarded over name and price.
- That direction suits how we have always bought. We have never chased labels for their own sake, and a list that rewards the estates doing the work in the vineyard is a list I am happy to put in front of our customers.
Anthony Ghattas, Sydney, September 2026
- Michelin press release: Bordeaux Grape Selection, 8 Sept 2026
- Michelin press release: Burgundy Grape Selection, 7 July 2026
- Michelin Guide: Bordeaux Wine Selection (full estate list)
- Michelin Guide: Bordeaux launch article
- Decanter: Michelin's statement on Latour
- Vino Joy News: a different Bordeaux hierarchy
- The Drinks Business: Burgundy launch
- The Drinks Business: Michelin launches Grapes rating