A vintage Champagne from a good house often costs two to three times its non-vintage sibling, and the explanation you usually get is that the year was exceptional. That is not an explanation. Plenty of exceptional years in other regions do not double the price of the wine. Something structural is going on, and it is worth understanding, because it also tells you exactly when the extra money is wasted.
Where the cost actually comes from
Three things, none of them romantic.
Time. The rules require non-vintage Champagne to spend a minimum of fifteen months in the cellar before release, twelve of them on the lees. Vintage Champagne requires thirty-six months. In practice serious houses go far past the minimum — six to ten years on lees is normal for a vintage bottling, and some prestige wines stay far longer. That is cellar space, insurance, capital sitting idle and stock losses, all multiplied across a decade.
Opportunity cost. This one is invisible and it is the largest. In a good year, the best base wines are the ones the chef de cave most wants for the non-vintage blend and for the reserve library. Declaring a vintage means pulling that material out of the system. The house then has to replace it in the blend, and it loses reserve depth for years afterwards. A vintage declaration is expensive before a single bottle is sold.
Scarcity by choice. Houses declare only when they judge the year good enough to stand without blending, so volumes are small and irregular. Some houses skip several years in a row. Low, unpredictable supply against steady global demand does what it always does to price.
Prestige cuvées — Dom Pérignon, Cristal, Bollinger La Grande Année, Krug's vintage, Taittinger Comtes de Champagne, Pol Roger's Sir Winston Churchill, Salon — stack a further layer on top: parcel selection, longer ageing again, and a brand that has been built for decades. You are paying for all of it, not only the year.
What the money actually buys in the glass
Not intensity. Not power. What a vintage bottling gives you is a wine with a shape — one year's acidity, ripeness and character showing through instead of being smoothed into a house profile — and the potential to develop over ten or twenty years into something with genuine tertiary character: toast, mushroom, dried fruit, iodine.
The word to hold onto is potential. It has to be realised, and realising it takes time in a cellar and attention at the table. That is the hinge on which the whole buying decision turns.
When it is not worth it
If you are drinking it young, cold, and at a party. A 2016 vintage opened in 2026, straight from the fridge, at a noisy dinner where the glass is refilled every ten minutes, will taste like a slightly leaner and firmer non-vintage. Every extra dollar went into a capacity for development that you are not letting the wine use. In that setting the house Brut is not the compromise. It is the correct wine.
If you cannot store it properly. This is the Hong Kong problem and it is not a small one. A bottle bought to keep for five years in a flat where the summer indoor temperature runs near 30 degrees and the air conditioning goes off when everyone leaves for work is a bottle being cooked slowly, then cooled, then cooked again. That cycling pushes the cork, lets in oxygen and pushes the wine into premature oxidation. You will open something that tastes old rather than mature, which is a different and much sadder flavour. Note that Hong Kong humidity is not the enemy here — 70 to 80 percent is genuinely good for corks. Heat and heat cycling are. If you do not have a wine fridge or a rented cellar space, buy vintage Champagne to drink within a year or two, not to hold.
If you are buying because a year is famous. Vintage charts describe the average of a region. Champagne is the region where averages mean least, because blending, parcel selection and press fraction give a producer more room to override the weather than almost anywhere else. A careful house in an awkward year regularly beats a mediocre one in a celebrated year. Buy the producer, then the year.
The claim that is simply wrong
That vintage is always a step up from non-vintage. It is not, and the counterexamples are famous. Krug's Grande Cuvée carries no vintage and outperforms most houses' vintage bottlings. Bollinger Special Cuvée, built on reserves held in magnum, does the same at its price. Charles Heidsieck's non-vintage Brut Réserve is aged far past what the rules demand. All three are multi-year blends, and all three would beat a great many single-year wines in a blind glass. Vintage is a different product, not a higher rank.
When it is genuinely worth paying for
When you want a wine that argues for one year, and you will serve it at 10 to 12 degrees in a white wine glass rather than a flute, with food that leaves room for it. When you are buying a birth year and the emotional value is the point. And when the house's vintage is a materially different wine rather than the same blend with a date — Bollinger's La Grande Année is fermented in barrel and behaves nothing like Special Cuvée.
There is one more case worth knowing. Late-disgorged releases — Bollinger R.D., Dom Pérignon P2, Krug Collection — have been held on lees by the producer for a decade or more in perfect conditions and disgorged only recently. In Hong Kong, where long-term private storage is the hardest part of the whole hobby, paying a producer to do that ageing for you is not a markup. It is buying the one thing your flat cannot provide.
Where to start
Most of the time the honest answer is that the house Brut is the correct wine and the vintage money is wasted. Buy a year only when you will give it a table, warmer glasses and an hour.
- All Non-Vintage Champagne — the right buy for a party, a fridge and a fast evening
- All Vintage Champagne — for the occasion where one year arguing for itself is the point
- Charles Heidsieck Brut Millésime 2018 — a house that ages far past the minimum and prints the dates, which is what you are actually paying for