How the rating works
The rating uses two things the NHS itself measures: how patients describe their practice in the 2026 GP Patient Survey, and how soon appointments take place, from NHS appointment data. We do not add our own opinions, and no practice can pay to change what is shown.
The two parts
The Practice score weighs two measures, each against its own England figure:
- Patient experience (70%): the share of patients who describe their practice as good. England: 76.7%.
- Appointments within two weeks (30%): the share of appointments that took place within 14 days of being booked, June to August 2026. England: 87.7%.
The weights are the same as on Vårdbetyg, our Swedish sister site, which weighs patient experience against the Swedish waiting-time guarantee in the same way. We did not choose the two measures ourselves: both are measures the NHS uses for GP practices.
Patient experience: the survey
The GP Patient Survey is run by Ipsos for NHS England. Each year it is sent to a random sample of registered patients aged 16 and over at every practice in England. Ipsos publishes, for each practice and question, the share of patients giving a positive answer, a 95% confidence interval and the number of people who answered. We show those figures as Ipsos publishes them. About the GP Patient Survey explains who is asked, how the answers are weighted and why every result has a margin of error.
We use one question: “Overall, how would you describe your experience of your GP practice?” (question 32), where “good” means the answers “very good” and “fairly good”. NHS England has used this question as its indicator of patient experience of GP services in the NHS Outcomes Framework, and it is the result Ipsos reports first for every practice. The question asks patients to weigh up their whole experience themselves, so we do not decide which parts of a visit matter most.
Appointments within two weeks
In Our plan for patients (2022) the government set the expectation that patients who need an appointment with their GP practice should get one within two weeks, and promised to publish, for every practice, how long people wait between booking an appointment and having it, so that patients can choose. NHS England publishes that data every month in Appointments in General Practice, and measured practices against the two weeks in its own indicator, ACC-08.
We count appointments the way that indicator does: the kinds of appointment patients usually want as soon as possible (general consultations, urgent appointments, triage, walk-in, home and care home visits), by any means (in person, by phone or online), whether the patient came or not. Planned clinics, procedures and medication reviews are left out, because they are booked ahead on purpose. The figure is the share of those appointments that took place within 14 days of booking, over June to August 2026, three months so that one unusual month does not decide it.
A practice needs at least 300 such appointments over the three months for this part to count. 18 rated practices do not have that many, or are not in the data; their score is patient experience alone, so missing data never lowers or raises a result.
What this figure does not show:
- NHS England publishes it as official statistics in development: the methods are still being worked on.
- It comes from practices’ booking systems, and practices record and categorise appointments in different ways. Practices whose system does not send the data are not in it.
- Time from booking is not the whole wait. A patient who could not get through to book is not counted, and a follow-up booked weeks ahead on purpose, at the patient’s or the doctor’s wish, counts as a long wait.
The Practice score
The Practice score is 70% of the practice’s patient experience figure plus 30% of its appointments figure, on a scale from 0 to 100. England’s score, worked out the same way, is 80.0. A practice is rated only when at least 30 people answered the survey question. Nothing is smoothed or adjusted, and the label is read off the same interval as the score.
On practice and area pages we also give the difference from England in points: the practice’s score minus England’s, with the interval moved the same way. It is the same figure said another way.
The margin of error
A survey asks some patients, not all of them, so every figure could be a little higher or lower if other patients had answered. The 95% interval shows how far. For patient experience we use the interval Ipsos publishes for the practice and question, which allows for how the survey weights the answers.
The appointment figure covers every recorded appointment, but a practice’s share still moves from month to month, with holidays, staff changes and how the diary is run. Its interval allows for that: it is worked out from the number of appointments, scaled by how much practices’ shares actually vary between months across England. The two parts come from different sources, so the score’s interval combines their margins as independent.
The other survey questions
Practice pages also show 12 more questions from the same survey, such as getting through on the phone, care and concern, and seeing the same GP, each with its own interval and number of answers and marked “Not part of our rating”. Until 22 October 2026 the survey part was the average of five questions. We changed it to the overall question alone because choosing and weighing the questions was our judgement, while the overall question is the measure NHS England uses and patients do the weighing. The same day we added appointments within two weeks.
The two labels
- Clearly above the England average: the whole interval lies above the England average.
- Cannot be told apart from the England average: the interval includes the England average. With the answers available, the survey cannot show a difference either way.
There is no third label. When the whole interval lies below the England average, the practice gets no label, and its figures are shown as for every other practice. We never put a practice under a negative heading or in a negative group. There is no list of the lowest-rated practices, for England or for any area.
Of the 6,163 open practices, 6,090 are rated: 1,666 clearly above the England average, 3,489 that cannot be told apart from it, and 935 with no label. 73 are not rated because they have too few answers or no survey result.
Every practice in one picture
The Practice score comes from a survey, which asks some patients and not all of them, and from appointment records, which move from month to month. So every score could come out a little higher or lower: that is the margin of error, and it is larger when fewer patients answer. At a typical practice about 110 patients answered the survey, and the score could be about 6 points higher or lower.
Each dot below is a rated practice, 6,090 in all. Further right means a smaller margin of error, usually because more patients answered. Higher up means the practice’s score is above England’s, lower down that it is below. The dots have no names.
The inner lines are where a practice’s difference from England is exactly as large as its margin of error. Between them, the difference could come down to which patients happened to answer, and that is where most practices are. The lines are far apart on the left, where the margin is wide, and close together on the right.
A dot above the upper inner line is clearly above the England average: its whole margin of error lies above England. It is the same rule as the label “Clearly above the England average” on a practice’s page, so the picture and the label always agree. The outer lines are stricter: 99.8% certainty instead of 95%, which takes a difference more than one and a half times the margin of error. Outside them, chance can almost be ruled out, upwards or downwards.
The lines show how certain a difference is, not how good the care is. The Practice score measures how patients experience their practice and how soon they get an appointment.
The large coloured dot is the practice whose page you came from.
Right now 1,666 practices lie above the upper inner line, 3,489 between the inner lines and 935 below the lower inner line.
The dots on the left spread out the most: with a wide margin of error, chance alone can move a result a long way. That is why most practices with a wide margin are “cannot be told apart from the England average”, whatever their score.
Areas
An area page covers one local authority district. The area’s result is the average of its rated practices’ results, each counting equally, with its own interval. We describe an area only when it has at least 3 rated practices. An area can be above, around or below the England average; that describes the area, never a single practice. Every practice in an area is also listed from A to Z.
Best-rated practices in an area
Each area page lists up to 10 of its best-rated practices, ordered by Practice score, with each score’s 95% interval and number of answers. Practices with exactly the same score share a place, and the name never decides a place. Only practices with a label are placed, so a practice whose whole interval lies below the England average never appears at the bottom of a ranked list; its figures are on the area’s A to Z list and its own page. An area needs at least 3 practices with a label to have a list. Neighbouring scores often overlap within their margins of error, so nearby places can be close calls.
Best-rated practices in England
The list for England places the 50 practices with the highest Practice score, by the same rules as the area lists, with one more: only practices labelled “clearly above the England average” are placed, so every place means the whole interval lies above England. We stop at 50 because further down the places are a tenth of a point apart and their margins of error overlap almost completely. Even within the list, the difference between two places is often within the margin of error. There is no list of the lowest-rated practices.
The awards
Each year’s survey gives two awards, the same two as on our Swedish sister site. They are free, shown on the practice’s page, and never for sale.
- Top 10% in England 2026: a Practice score in the highest 10% of all rated practices in England (90.9 or more this year), and labelled clearly above the England average. Practices level with the cutoff all get it. 581 practices hold it.
- Top-rated in an area 2026: first place on the area’s list of best-rated practices, for a local authority area or for London. An area needs at least 3 practices with a label, so first of two never counts. Practices with exactly the same score share first place and all get it. 297 practices hold it.
A practice needs at least 30 answers for an award. Practices for one group of people and services you cannot choose never get one. An award belongs to the 2026 survey and stands until the next one is published. A practice can license a graphic of its award; the licence never changes the award, and a practice cannot buy one it has not earned. For practices.
Nearby practices with clearly higher results
On a practice’s page we list up to two practices within 10 km (as the crow flies) whose result is higher by at least 2 points and by more than the combined margin of error of the two practices. We list only practices that have a label at least as good as the practice’s own. Registration depends on where you live: most practices only accept patients inside their catchment area.
Services that are not ordinary practices
NHS England’s register of organisations lists some services as GP practices although most people cannot register with them, and it gives them the same codes as any other practice. So we check them by hand: we look for them by name, by the size of their list and by the age and sex of their registered patients, and then read what the service itself or the NHS website says it is for. Each one is recorded with the page we checked.
- 41 practices register only one group of people, such as people who are homeless, asylum seekers, students or care home residents. Their pages and ratings stay, with a line saying who they are for, and they are on the map and in the postcode search with that line. They are never placed on a list of best-rated practices or suggested as a nearby practice with clearly higher results.
- 45 are services you cannot choose to register with, such as special allocation schemes (where patients are placed by the NHS), walk-in and urgent treatment centres without a list, and extended-hours hubs. They keep a page that says what they are, but they are not on the map, in the postcode search, in the area lists or on any list of best-rated practices.
What the rating leaves out
- The Care Quality Commission’s rating. We show it beside our rating as a separate fact; it is never part of it.
- The survey’s other questions. Practice pages show 12 of them as facts beside the rating.
- Clinical outcomes and staffing. Practice pages show some of these, such as vaccinations and the number of GPs, as facts beside the rating.
- The other appointment figures, such as same-day appointments and how many are face to face. Practice pages show them as facts.
- Anything a practice tells us, and anything anyone pays for.
Every data source and licence is listed on Sources and licences.
Sources
- 2026 GP Patient Survey: https://www.gp-patient.co.uk/
- Contains information from NHS England, licenced under the current version of the Open Government Licence