The number of votes cast by viewers of the Eurovision Song Contest is unknown. Each year, the data revealed after the Grand Final from the European Broadcasting Union shows which countries received more votes than others from each participating nation, ranking them from first place to last place, but there is no indication given of how many votes each song received or the amount of votes cast in the country.
Historical records show some idea about the number of votes cast at the Song Contest. Reports from ESCToday in 2005 claim that in the ten-minute voting window during Kyiv 2025 there were over 5 million votes around the continent. A BBC factsheet for press before the 2012 Contest in Baku reported that the 2009 Contest saw 10 million votes cast across Europe, as Alexander Rybak took a landslide victory.
Since then, there has been no data available about the number of voters at the Eurovision Song Contest. Until last week. As the doors closed on the 2025 Song Contest in Basel, Belgian broadcaster VRT and Spanish broadcaster RTVE revealed data on the number of voters they have had at Eurovision this year (and the past two years in the case of the Belgian broadcaster).
With these two sets of data (and I will say now, an awful lot of assumptions), let us make a model, one attempting to predict how many votes were cast at the 2025 Eurovision Song Contest.
The Number of Votes at Eurovision
The data from the Belgian broadcaster tells us there were 220,554 voters at this year’s Eurovision Song Contest Grand Final voting from Belgium. This is a number that is higher than the year before, where the number of voters was just over 200,000, and significantly higher than in 2023, when nearly 130,000 votes were cast.
These numbers from Belgium are more than those who voted in Spain this year. In Spain, there were 142,688 votes cast during the Grand Final of Eurovision 2025. Like in Belgium, the majority of those votes were cast online, via the esc.vote system.
With the delegations of these two nations revealing their voting numbers, we have decided to try to build a model to work out how many votes have been cast around Europe. As one can imagine, a lot of guesswork and assumptions are involved. Many different factors can impact the number of voters from a country at the Eurovision Song Contest, including…
- …The cost of voting in each country
- …The amount of disposable income residents have
- …The engagement that residents have in Eurovision, and in voting in the show
- …The number of viewers watching in each country
Of course, there are many other reasons in addition to these we pick out, such as the fame of certain artists in the country, engaged diaspora groups, and even the time zone the country lies in. However, what we have been able to do is create a model based on these factors that can spit out the Belgian and Spanish televote numbers very close to the recorded amount.
Our model notes that in Belgium, televoting costs are lower than in Spain (€0.75 in Belgium rather than €1.10 in Spain) and that Belgian residents have a higher average net income than in Spain. Furthermore, we know through voting data at Benidorm Fest that Spanish televoting numbers are not exceptionally high (just 17,000 votes in the final this year, although an app vote was also in place). As such, there is evidence that the Spanish public is less likely to participate in Eurovision voting than other parts of Europe. Therefore, despite the larger population and viewing figures, these factors result in Spain having fewer active voters than Belgium at the Song Contest.
If we apply this model to the other countries, we see a considerable spread in the number of votes cast from country to country. Unsurprisingly, with the largest viewing figures and cheap voting rates, our model anticipates that the German and British voting public have the most significant number of voters, with our model estimating that within each nation, there are just over four million votes cast.
Our models top 10 projected nations for number of votes at Eurovision, with viewing figures provided by Eurovoix.
On the flip side, we believe some of the smallest televoting numbers to be in some of the continent’s easternmost fringes. Just 47,000 viewers watched the 2024 Eurovision Song Contest from Azerbaijan, where the show begins at midnight local time. From those viewing figures, coupled with Azerbaijan’s relatively low average income, our model spits out that they can expect just 1,200 votes to be cast in Azerbaijan in the Eurovision Grand Final. (Yes, our model does use 2024 viewing figure data as much of the 2025 data is not available as of time of publication This also means that we ignore any possible impacts on viewing figures on if a country qualifies to the Grand Final or not—and Azerbaijan did not last year—for this modelling).
The ten lowest projected televotes according to our model, alongside viewing figures provided by Eurovoix
All combined, these estimates are based on data from Spain and Belgium, which take us to a total number of votes in the Grand Final of around 17.2 million. Given all the guesstimates and extrapolations made, I would not read much into the exactness of this number. Still, I am relatively confident that our final figure is closer to 17.2 million than 1.72 million or 172 million. This model obviously excludes any votes from the new Rest of the World vote, which can vote before the Contest takes place and is less linked, therefore, to direct viewing figures.
We Need More Data
I will be the first to suggest that these numbers are almost certainly huge overestimates at the high end and huge underestimates at the lower end. We don’t have much data from the modern day Song Contest to compare, but we know there were 323,000 votes cast from the UK in 1999, and that the UK selection show of 2005 saw this number go up to around one million voters. Yes, voting audiences are more used to televoting today and in many ways the Song Contest and the engagement with the Contest is higher today, but to expect ten times the votes at Eurovision today compared to a generation ago is hyperbolic.
That said, there are legitimate reasons why the number of voters might be higher now than previously. It is now possible to vote throughout the broadcast of the Grand Final and the esc.vote mechanism is much more streamlined, especially if voting numerous times for one entry, compared to sending 20 different phone calls or text messages.
While Azerbaijan and other nations have a particularly low projected number of votes, our model assumes a linear relationship for voting prevalence from both income and voting cost, which fit the data from Spain and Belgium. However, it is likely that this relationship is not directly linear (making the voting 50% cheaper does not result in 50% more votes) so this may exaggerate the lower end of the scale, particularly in those poorer nations with lower incomes. We have assumed that a 50% lower income means a 50% lower vote ratio for this modelling, which we know is not accurate (rich people are not the only ones voting twenty times).
With more data, we would likely be able to move this linear model away to one that is more nuanced and could be a better representation. Spain’s televoting figure is much lower than Belgium’s, not because of income disparity or voting cost, these are in reality only factors of slight variance, but instead it is cultural within Spain which is not a nation with large televoting traditions.
The number our model spits out, around 17 million voters, is a sensible ballpark figure given what we know about the modern Eurovision Song Contest and how its reach has expanded in recent years. But if we were going to actually achieve a reasonable prediction on the number of voters at the Song Contest, we do need more data from a wide range of participating broadcasters to be made available.
However, the reveal of such data from Belgium and Spain this week also makes us consider something else.
Why isn’t this data normally made available? What is it that our organisers and broadcasters have felt for the last twenty years that the public should be shielded from? Our Eurovision Song Contest, bringing together millions of people from around the world each year, should be the pinnacle for not just song and spectacle but also for transparency and openness. There are National Finals around Europe that reveal the number of voters even before the credits start to roll. Why doesn’t the biggest of them all do the same?