Every summer the same number goes around. At 20°C, Britain hits its “rosé tipping point” and sales of pink wine jump by 150%. Push past 26°C and rosé apparently overtakes red and white altogether. It gets repeated in the trade press, in newspapers, in wine newsletters, and by people at parties who are very pleased to know it.

We went looking for the research behind it this week. There isn’t any.

Where the number actually comes from

The 20°C figure comes from a Waitrose press release issued on 16 June 2025, headlined “Rosé sales surge by 150% when 20°C hits”. That is the entire source. There is no methodology, no sample size, no stated time period, and no named analyst. The only person quoted in it, Nicki Hobbs, is Waitrose’s rosé buyer, and she is talking about serving temperature, not about the 150% figure.

The 26°C figure is a different company in a different year. Majestic Wine told the trade press in July 2022 that rosé would overtake white and red above 26°C, based on a single week of its own sales from 26 May to 1 June 2020. Again, no methodology.

So when you see the tidy “between 20°C and 26°C, drinkers move to rosé” framing, you are looking at two different retailers’ marketing, from two different years, based on two different datasets, spliced into one sentence that neither of them wrote. It has been laundered into a fact by repetition.

We want to be fair here. Neither company did anything wrong. Retailers put out seasonal sales notes all the time, journalists pick them up, and nobody involved claimed to be publishing science. The problem is what happens two years later, when the number has lost its footnotes and is being quoted as though somebody ran a regression.

What the actual research found

There is real work on this, and it is more interesting than the press releases.

Martin Hirche, Juliane Haensch and Larry Lockshin, at the Ehrenberg-Bass Institute, published a study in the International Journal of Wine Business Research in 2021. They took weekly retail sales from 37,346 stores across 651 US counties over three years and modelled them against daily temperature. This is not a press release. This is a proper time-series analysis on a very large dataset.

Their finding, in their own summary: beer, liquor, red wine and white wine are temperature sensitive throughout the year. Rosé, sparkling and other wines are not.

Read that again, because it is the exact opposite of the thing everybody repeats. In the largest study of its kind, rosé was one of the categories that did not track the thermometer. The authors were candid that this ran against their own expectations.

There was something that moved sales hard across every single category, though. Not weather. Public holidays. Easter, Thanksgiving, Christmas, New Year. It turns out people drink when they have time off, which is either a profound insight into human nature or the most obvious sentence ever written, depending on your mood.

The honest caveats

We are not going to overclaim in the other direction, because that would be doing the same thing in reverse.

That study is US off-trade data from 2013 to 2015. Britain is not America, supermarket rosé culture in the UK is genuinely its own thing, and a decade has passed. It is entirely possible that a well-designed UK study would find a real temperature effect on rosé.

It is also true that retailers see real spikes. When Majestic says it sold a bottle of rosé every 12 seconds during a heatwave week, we believe them. But a busy week during a hot spell is not the same claim as “20°C is the threshold”. Hot weather in Britain arrives in June, alongside long evenings, bank holidays, barbecues and the start of the summer social calendar. Untangling the temperature from everything else it arrives with is exactly the hard part, and it is exactly the part a press release does not do.

The honest position is this: nobody has published credible evidence of a rosé temperature threshold in the UK. The one large study that looked properly found no temperature effect on rosé at all. That is where the evidence sits today.

Why we care about this at all

Because this is the whole game, really. Wine is full of numbers that sound authoritative and turn out to be somebody’s marketing, or somebody’s guess, or a real figure that got separated from its context three retellings ago. The 100-point score is the big one. The “rosé tipping point” is a smaller, sillier cousin.

None of this means you should stop drinking rosé when it is hot. Cold pink wine on a warm evening is one of the genuinely great simple pleasures, and you do not need a supermarket to hand you a statistic as permission. Drink it in February if you want. The wine will not check the forecast.

The temperature number that is actually worth knowing

Here is the irony. There is a genuinely useful temperature fact in this story, and it is the one nobody repeats: the temperature of the wine itself, not the air.

Jancis Robinson calls serving temperature the single most important aspect of serving wine, and she is blunt that the old rule is dead. “The old ‘room temperature’ ideal serving temperature for red wines belongs to a distant, chillier age,” she writes, noting that most reds get served too warm and most whites so cold you cannot smell them.

The WSET makes the same point with numbers. “Room temperature” for wine means 15 to 18°C, not the 20°C your living room actually sits at. Reds above 18°C lose their freshness and their structure goes muddled. Chill anything below 6°C and you flatten its flavour entirely.

Rough guide, if you want one:

  • Rosé and light whites: 7 to 10°C. Fuller rosés, like a proper Tavel, are better nearer 12 to 13°C.
  • Full-bodied whites: 10 to 13°C. Straight from the fridge is usually too cold.
  • Light reds (Gamay, Beaujolais, simple Pinot Noir, Cinsault): 10 to 14°C. Yes, that means from the fridge. Do it.
  • Full-bodied reds: 16 to 18°C, and no warmer.

If you take one thing from this piece, make it that: half an hour in the fridge does more for a light red than any amount of knowing what temperature Britain allegedly starts drinking rosé at. We have written more on red wine serving temperature and on rosé itself if you want to go deeper, and a few more of these in our five myths about wine.

“I love that the real finding is so much better than the fake one. Everybody wants a neat rule about when to drink rosé. The research says the rule is ‘when you are off work’, which is the least glamorous and most human answer possible. I’ll take it.” — Antoine

So no, there is no rosé tipping point. There is just hot weather, time off, and a drink that happens to be very good in both. Chill it properly and stop worrying about the thermometer outside.

Sources: Waitrose / John Lewis Partnership, Drinks Retailing, International Journal of Wine Business Research, WSET