The Church of England has released its Cathedral Statistics 2025, and an accompanying press release which is copied below.
Attendance at Cathedral services rises for fifth year running
31/07/2026
A total of 32,600 people attended weekly services – such as Evensong, Morning and Evening Prayer or Sunday Eucharist – at Church of England cathedrals in 2025. Sunday attendance was up notably among adults, at eight per cent higher than in 2024.
There was also a six per cent overall increase in Christmas attendance and five per cent rise in Easter attendance.
Additionally, 657,700 people attended 2,080 specially arranged services over the year and 295,000 at additional regular services such as services for schools.
The figures show the number of choirs in Cathedrals rose to a record high of 208, mainly due to a big rise over the last decade of mixed choirs.
Cathedrals also hosted 1.83 million people at public or civic events, such as graduation ceremonies, and welcomed 9.23 million visitors.
Although attendance, participation, and engagement was, by many measures, higher in 2025 than in 2024, it was lower than in 2019 overall. However, Easter attendance drew level with pre-pandemic levels for the first time.
The Dean of Wakefield, Dr Philip Hobday, said: “Cathedrals are a cool space in the heatwave, a warm space in the winter, and in some of the nation’s most disadvantaged communities, one of the few spaces which are free to enter and where all are welcome.
“We are thrilled to see a wider range of people of different ages and backgrounds coming into our building, whether to attend a special Christmas or Easter Service or to enjoy music concerts in all styles from Bach to Bowie, to support a local up-and-coming artist or a school event, or just for a few moments of peace in troubled times.
“Across our cathedrals, over 208 choral groups keep a distinctive rhythm of daily worship alive, and Christmas and Easter services have seen noticeable increases in attendance.
“With no guaranteed public sector funding, English cathedrals rely on the hard work of staff and volunteers, and the generosity of personal givers and funding bodies, to keep these beautiful spaces open to all as places of heritage, community-building, and Christian witness.”
Sorry to be a doom merchant, but i think the spin of a year on year increase is taken massively out of an overall context of decline. Numbers are nowhere near pre-pandemic numbers – and the Community Roll is the most telling red flag of working with an ever shrinking pool. That has plummeted since 2019. Interestingly, the place of music is encouraging, and the rise of choirs is a good sign, but taken over a long period of time – the numbers do not look good. I wish we were more honest with statistics.
The compilers of this report (led by the outstanding Ken Eames) are completely honest and put many other statistical exercises to shame. There is a great deal of transparent detail, extending even to the treatment of missing data and a copy of the actual return.
However, the Department of Cherry Picking has been at work in writing the press release, as usual.
Thank you for your kind words, Bernard, which are more generous than I deserve.
I don’t think this is the right setting for me to offer any thoughts I might have about press releases, but I will say that I always find my colleagues in the Comms team (whose job is far more stressful than mine) to be generous with their time, and I’m grateful for their help in making people aware of the cathedral statistics report.
Thanks Ken—-I think this is the nub of it. If you think about the way that ONS is supposed to work, the production and publication of the stats is independent of the policy use. So ONS produces its own press releases and only later can the government react to them. Indeed during my time working in Government there was a big tussle about “pre-release access” which was, as you will know, eventually abolished, so ministers and civil servants only see the stats when they are released to everyone. But what is going on with your work is that the comms… Read more »
Ken Eames as ever has been enormously professional and has produced a report which is scrupulously transparent and informative. However, as usual, the Department of Cherry Picking has been doing its thing. For fun, I loaded the report into ChatGPT and asked it to produce an honest press release. Then I loaded the actual press release and asked for a comparison, which is summarised as follows: (Here, “I” is used anthropomorphically to mean ChatGPT).. I’d call the actual release acceptable but advocacy-led. It is not egregious because it contains the 2019 caveat, but it foregrounds every positive figure and omits… Read more »
Whatever the merits or otherwise of this exercise, is using an AI tool defensible, given the huge damage that this technology does to the environment? Or is net zero yesterday’s concern?
Using an AI tool to expose authors of dodgy press releases is as indefensible as driving round in a car or getting a flight instead of the train.
Homeless Anglican, I’ve tried in the report to give the comparison with 2024 and 2019 figures approximately equal prominence (since I think both are useful), as well as including the usual charts and tables showing a more detailed time series. I’m sorry if any of that comes across as dishonest; that is not my intent. I’m never sure whether I’m giving way too much information in these reports, but if there are things that you or anyone else think ought to be added please tell me. I am open to suggestions.
Ken – forgive me. My intention was not to say that the numbers were dishonest – they look assiduously collected and dissected. I just felt that the reporting of them and the headline management/comms/ dept of cherry picking has highlighted a particular aspect to report – rather than a wider more discerning overview of the work you have done. I wanted to separate the report (good!) from the reporting (bad!). I hope that clarifies things! 🙂
A quick question: who does the counting? Is there someone at every service tasked with doing an accurate headcount and then filling in the register? I’d have thought that, if numbers are merely estimates, there is a strong tendency to err on the generous side.
I can only speak for Southwell Minster where the counting is done assiduously by the vergers.
Andrew, that’s a very reasonable question. I’ve asked my cathedral colleagues how they do their counting, and I agree with Laurence that it’ll often be vergers/sidespeople/stewards. Quite possibly with clickers. It probably varies from place to place and from service to service. Realistically, even the most assiduous approach will involve some estimation on occasions. And there are always questions such as whether or not to “count” someone who arrives late, sits at the back, and leaves early. If there were a tendency to under or over count then unless that tendency itself has changed over time the trends would still… Read more »
When i was in cathedral ministry, the view was that Deans / vergers tended to underestimate attendance. Of course it is very difficult to substantiate that. However I would say that as a Dean I never had any role in counting any attendance figures and I suspect that is pretty standard. It was always the job of the verger.
Smaller churches have an advantage here. MY last parish in Edmonton (western Canada) had an average Sunday attendance of around 80. After I had been there about five years (I stayed almost 24), I made a habit of running down the parish list every Sunday after church and just noting down who had been present (the list made this easy); I always kept a rough count in my head of people I didn’t know, and then added them to the total parishioners present. My observation was that three times out of four, the sidespeople had undercounted the attendance.
“Cathedrals are …one of the few spaces which are free to enter and where all are welcome.” Is it universally true that cathedrals are free to enter?
During worship, yes, at other times not so much.
Technically – yes – to go and pray or to a service, but to sightsee is a different thing altogether. I was once refused entry to a service in York Minster because they didn’t believe that i wanted to worship rather than buy a Minster Tea towel!!
My daughter and I were only allowed by a steward to enter Coventry Cathedral to attend evensong on the condition that we didn’t look at anything.
They don’t charge for services.
Surely, Most cathedrals charge for admission unless your attending a service.
The four nearest to me, Birmingham, Lichfield, Coventry and Worcester do not charge for admission.
Nor does Southwell
No , that’s incorrect. I don’t know how many charge but quite a few. Even if you say you have come to pray you may be ushered into a specific place and not be able to access the rest of the building.
I think currently around 8 cathedrals charge for entry. None of the formerly “parish church” cathedrals have that option!
When you say they don’t have that option, is that hust because they dont hace the same visitor numbers or the same maintenance costs. Three of the 19th century creations, which up till then had functioned as parish churches, Ripon, Southwell and St Alban’s, are very much the equal of older cathedrals in architectural prominence and historical interest, and probably attract similar numbers of visitors. Two others, Manchester and Newcastle, are very large mediaeval buildings with, I would guess, significantly high maintenance costs. Is there anything stopping any of these charging for admission if it was believed it would help… Read more »
“Just over half of the cathedrals reported having some paying or donating visitors.” – page 16.
I don’t have at my fingertips the number that reported both paying and non-paying visitors, but I can find out if people would be interested.
Could it be that increasing (recovering?) attendance midweek is largely elderly/ tourists on a day out (with Senior railcard, free bus pass?)- changing/ unchanging demographic?
Yes, it could be. I don’t have evidence one way or the other on that. If so, it might have been the case pre-pandemic too.
Thank you Ken for this, and other replies. I admire your tenacity and thoroughness, and join Bernard in his views. The underlying question for me is what, if anything the figures mean, and what if anything should happen in response- put rather crudely ‘so what?’ I doubt there is much if any evidence as to age, ‘status’ etc of visitors, and perhaps for the more Evangelical amongst us whether or not they are, or become, ‘regular communicants’ having ‘met with God’. the nation has an ‘ageing demographic’ as does ‘the church’- is there a correlation? Thank you for all you… Read more »
The totals are only estimates as two cathedrals did not submit returns in 2025. So there we have it lies, damned lies and statistics!
No, lies, damned lies, poor statistics, and AI.
what is your evidence for use of AI?
No AI was used at any point in any of the work relating to this report. All errors are human errors, most likely mine.
AI has its uses; I gather for instance it’s good at coming up with a helpful and tweakable starting point for code to do a particular task. I’ve not myself used it yet.
Yes, it does. Started using it in that context about a month ago – I’ve been surprised. I am currently using it to create boiler plate code for modelling waste processing plants.
My original comment re. AI was simply modernising an old saying – a joke.99% of statistics I see is very bad statistics. 1% is useful.
Thanks Nigel. Good luck finding the 1%. There’s another stats joke, isn’t there: 67% of statistics are just made up on the spot.
This exercise is one of the most complete data collections (I won’t use the word survey because that suggests sampling rather than exhaustive collection) I have ever seen. It’s really good work, including Ken’s top line report which makes clear how gaps and missing data are dealt with. It involves a lot of effort on behalf of a lot of people. The only time I’d be slightly cautious is when a count is based on “every seat is full” because in practice that may not actually be so, but this would be a very small effect. And as pointed out… Read more »
Why don’t we just stick to the clickers? Burning a bit more of the planet to spare ourselves the effort of counting people doesn’t strike me as the brightest idea…
I agree, just use clickers. We use them to count communicants and for October count. Works perfectly well.
Rather than counting people during the service, in which some may be eclipsed by larger pewmates; the Hanrahan approach can be used. Count them all in and count them all out.
Clickers seem to be pretty effective, and especially handy when there are several doors. I like the “Hanrahan approach” too. I suspect I mess up either system on occasions by going in and out with small children. It’s inherently harder to count accurately if there are people attending worship and non-worshipping visitors in the building at the same time. More relevant to cathedrals than most churches, though of course there are some parish churches that also welcome large numbers of non-worshipping tourists/visitors. My instinct is that the people doing the counting know what they are doing and generally report numbers… Read more »
A further problem when filling in attendance figures does not seem to have been mentioned This is distinguishing children (defined as under 16) from adults. “Excuse me sir but is your daughter over 16” may be met with a rather robust response. Asking the young lady directly, even more so.
In October, as required by the Church of England, we count all who come into the building for each service. As also required by the Church of England we count all communicants at every communion service. Clickers are the best way of doing this as there are often over 300 attending a service in October and over 250 communicants.
Adrian, yes, as it says in the report I estimated figures for two cathedrals this time around. I take a fairly simple approach – outlined in the methodology. I’d be very surprised if the presence of estimates makes much difference, though it’s possible that something extra-ordinary happened in one of those cathedrals – a surge in adult baptisms, or all Christmas services cancelled because the heating broke down, or something like that. Back in pre-pandemic days we used to chivvy for months until we had figures from every cathedral, which was fine in a way but slowed down publication considerably.… Read more »
You should talk to my niece Mary Gregory (ne Atwell), who is currently.Executive Director for Population, Census and Social Statistics at the ONS. Her father James Atwell was dean of Winchester.
Thank you for posting the link to this report, and thank you to everyone who has commented. I’ll try to answer a few queries in the thread. As ever, I’ll be delighted to receive constructive suggestions for improvements, or to have detailed conversations about methodology – do get in touch using the contact details included in the report if you’d like to (though I’m on leave at the moment, so apologies in advance for slow replies).
Hi Ken
I have a doctorate of theology and ministry from Durham as well as a theology MA from KCL. Happy to be involved pro bono for anything you think I might be able to help with. I am not a professional academic, just a retired ex-teacher and HR managerwho needs to exercise his brain!
In general Cathedral services are much easier to allow people to be anonymous and not, as in so many smaller churches, be seen as another potential volunteer for the coffee rota. The downside, as in my large parish church, is that it gets harder to know people and so offer – not dragoon – into some greater involvement / commitment. I wish that I knew the answer so that the same oldies like me, can be reliably replaced – or is 21st century idea of commitment different from that in which I was taught in the 1960 /70s?
Some developing hypotheses here? Former (lapsed?)’churchgoers seeking a kind of sanctuary, freedom from being prevailed upon and perceived ‘guilt’ at not being an active ‘servant’ but able to just receive, and ‘be there’ in a transcendental atmosphere? Young folk, busy, phone attached to the ear 24/7, ‘free’ of commitment (and of meaning, purpose). Perhaps these are our perceived pre-conceptions? Ken’s stats have some potential for data, but will still need to be attended to in terms of what that means for ‘church leaders’. I imagine that to an extent at least Ken’s time and priorities are directed (guided?) by ‘the… Read more »
It is worth reading all the recommendations for the NST . Maybe not the most riveting of reading if you are not interested in the nuts and bolts of running consistent and sound safeguarding processes but what jumped out to me was just how basic some of them are….
That these processes are missing should be a source of deep embarrassment to the powers that are and especially those seeking to block professional progress
From my own experience in a large parish church several things help address your concerns. Creating a welcoming atmosphere in which people talk to each other. Encouraging all who attend to join a small group so that everyone can get to know others better. Finally, personal invites to join an area of service, e.g. coffee rota, gardening group, etc. Finally, some do just want to attend services, and then go.
People may go to Cathedral services because they appreciate being in a beautiful space with beautiful music. It does not necessarily have any implications for their adherence to the Christian faith. On the other hand, someone with a deep adherence to the Christian faith might sit and worship beside a river. Or on the top of a mountain.