NHS GP appointments: what the latest figures actually show

England’s general practice system is not short of appointments in the headline sense. NHS England Digital’s May 2026 bulletin recorded 30.0 million appointments across general practice, and 44.9% of them took place on the same day they were booked. Those figures describe completed activity in appointment books — not how hard it was to get through on the phone at 8 a.m., and not whether a practice’s evening line was answered. Reading the two layers apart is the difference between a useful access conversation and a misleading one.

May 2026 is the latest official month — and the numbers are large

The Appointments in General Practice, May 2026 bulletin, published on 25 June 2026, is the current official read from NHS England Digital for England. Its key facts are precise enough to quote without rounding theatre:

• 30.0 million appointments were recorded across general practice. Of these, 29.0 million (including Covid-19 vaccinations) were recorded in general practice appointment books from 98.9% of GP practices.
• 44.9% of appointments in May 2026 took place on the same day that they were booked.
• 60.5% of all appointments in May 2026 were carried out face to face.
• 89.6% of all appointments were attended.
• 44.5% of all appointments were carried out by a GP, and 18.4% by nurses.

That last split matters for anyone still picturing “a GP appointment” as synonymous with “an appointment in general practice.” A substantial share of recorded activity is delivered by the wider practice team. Capacity conversations that only count GP-delivered slots will understate what the books actually held in May — and overstate how much of the 30.0 million figure was one profession’s diary.

The series landing page at NHS England Digital publishes these bulletins monthly. As of late July 2026, June 2026 data was still listed among upcoming releases — so May remains the latest month a reader can cite from the official series without reaching for an unpublished figure.

RCGP’s February summary is consistent — not identical

The Royal College of General Practitioners maintains a key statistics and insights page that restates NHS Digital series figures for public readers. Its February 2026 monthly block is useful as a second, earlier snapshot — and dangerous if someone averages it with May into a fake “spring 2026” composite.

For February 2026, RCGP reports that general practice and primary care networks delivered approximately 31.8 million consultations, including Covid vaccinations; that approximately 14.1 million consultations were delivered on the same day as booking (44.4% of all general practice appointments); that around 26.4 million consultations were delivered within two weeks of booking (83.1%); that approximately 19.5 million appointments were face to face (61.2%); and that approximately 14.2 million appointments were delivered by a GP (around 44–45%).

Same-day access in the mid-44% range appears in both months. Totals differ because the months differ — February’s 31.8 million is not a correction of May’s 30.0 million, and May’s 44.9% same-day share is not “up from” February’s 44.4% in any causal sense without a full time series in hand. The honest move is side-by-side citation, not a blended headline.

Volume is not the same thing as getting through

Appointment counts answer a specific question: how much activity was recorded in practice books, how soon after booking it happened, and in what mode. They do not answer whether a patient spent twenty minutes in a phone queue at 8:01, abandoned the call, walked into an urgent treatment centre, or tried again the next day. A practice can post strong same-day percentages in the national extract while still running a phone line that fails the first hour of the morning — because the people who never get an appointment booked never enter the denominator.

That distinction is the operational parallel to what US practices see when after-hours voicemail quietly taxes the next day’s schedule. The US version of that argument — measuring unanswered evening demand as an invisible cost — is set out in after-hours calls as a practice cost. England’s national statistics are richer on completed appointments than most US public series; they are not a substitute for practice-level telephony metrics such as abandonment, average speed to answer, and after-hours disposition.

Attendance in May was high at 89.6%, which means the complementary share — patients who did not attend — is still large enough in absolute terms, on a 30-million base, to matter to any practice watching empty slots. Reminder and confirmation discipline is part of protecting that attendance. Practices that automate those messages still need clear consent and content rules; the US TCPA framing in appointment-reminder compliance is a different legal regime, but the operational lesson travels: keep confirmation channels narrow, countable, and auditable. For a rough sense of empty-slot cost, practices can also use the no-show calculator once they have a local DNA rate — the national 89.6% attendance figure is a system average, not a practice target.

What the workforce pressure around the numbers tells you

RCGP’s broader key-statistics page also surfaces structural strain that appointment totals alone cannot resolve. Practices in areas with the highest income deprivation had, as of October 2022, on average 300 more patients per fully qualified GP (14.4%) than practices in the least deprived areas — a discrepancy the College notes had increased 50% since October 2013. A 2025 GP Voice Survey of 2,316 GPs cited on the same page found that 73% of GPs say patient safety is being compromised by excessive workload. Those are not appointment-book statistics; they are context for why “just add more appointments” is an incomplete response to access complaints.

For a practice manager in NHS England, the actionable read of May 2026 is therefore layered:

1. National activity is high — 30.0 million appointments in a month is not a system that stopped seeing people.
2. Same-day remains a large share — 44.9% is a structural feature of how demand is currently met, not a rounding error.
3. Mode and who delivers care have shifted — face-to-face is still the majority at 60.5%, but GPs delivered under half of recorded appointments.
4. Local phone access is a separate dashboard — if the practice only watches the national series, it will miss the calls that never become appointments.

What practices and commissioners can do with this

England’s appointment series is one of the clearest public windows onto primary-care activity in the English-speaking world. Used carefully, it shows a system still delivering tens of millions of contacts a month, nearly half of them same-day. Used carelessly, it becomes a shrug that “access is fine” — aimed at patients who never got their call answered. For the wider English-language product context around scheduling and confirmation, start from Arbol’s English home.