Exam fundamentals11 min read·25 July 2026

PLAB 1 Pass Rate and Pass Mark: What the Numbers Actually Mean

The figures online contradict each other — pass marks quoted from 108 to 126 out of 180, pass rates from 50% to 75% — and most are published with no source at all. How the pass mark is really set (Angoff plus one standard error), why the figures legitimately differ, and why the pass rate can't predict your result.

Revising this topic? See where you stand first.10 free AKT questions — no account

There is no fixed PLAB 1 pass mark, and the pass rate changes with every sitting. If you've been searching for both numbers and getting different answers everywhere, that's not you being unlucky with sources — the figures circulating online genuinely contradict each other, and most of them are published without any citation at all.

We went looking properly. Here is what's verifiable, what isn't, why the numbers disagree, and — more usefully — why the pass rate is close to useless for predicting your own result.

The contradiction, laid out
How the pass mark is actually set
Where the real data lives
Why published figures legitimately differ
Why the pass rate doesn't predict your result
What actually predicts it
If you fail: the attempt rules
FAQ

1. The contradiction, laid out

These are figures we found quoted in third-party PLAB guides, all presented as fact, all for the same exam:

Quoted "pass marks" (out of 180):

108–117
110–118
115–120
116–126

Quoted "pass rates":

"around 65%"
"60–75% on first attempt"
"roughly 50% for IMGs on the first try"
"62.2% in 2025"
"70.5% overall in 2024, 75% first attempt"
"55–75% depending on the sitting"

Those cannot all be right. The pass-mark range spans 18 marks — a full 10% of the paper. The pass-rate figures span 25 percentage points.

And when we checked the article currently ranking well for "PLAB 1 pass rate", it made broad statistical claims — a 55–75% historical range, a "typically around 60–65%" pass mark — without attributing any of them to a source. That's how these numbers propagate: one unsourced claim gets quoted by the next guide, which gets quoted by the next, and eventually the number looks like consensus because you've seen it five times.

So treat any single figure you're shown — including in this post — as needing a source. Below is what actually holds up.

2. How the pass mark is actually set

This part is verifiable and it's the most useful thing on this page.

The GMC sets the pass mark for each PLAB 1 paper using the Angoff method, and includes one standard error of measurement. That's the GMC's own published description of its process.

Unpacked:

The Angoff method is criterion-referenced. A panel of experts reviews each question and judges what proportion of minimally competent candidates — a doctor at the required standard — would answer it correctly. Those judgements are aggregated into a pass mark for that specific paper. The standard is defined by what a competent doctor should know, not by how your cohort performed.

So the pass mark moves with the paper. A harder paper gets a lower pass mark; an easier paper gets a higher one. That's the entire purpose: it holds the standard constant while the questions vary. Which is precisely why no fixed number exists to quote, and why every guide confidently naming one is misleading you.

One standard error of measurement is added. Every test has measurement error. Adding one SEM is a deliberate conservatism, and it means the operational threshold sits slightly above the raw Angoff figure.

Three things follow, and they matter:

There is no quota and no curve. You are not competing against other candidates for a fixed number of passes. If everyone sitting your paper meets the standard, everyone can pass. Nothing you gain depends on someone else failing.
A "hard paper" is not bad luck. If your paper was harder, its pass mark was set lower. The difficulty is priced in.
Aiming at a threshold is the wrong strategy, because you don't know your paper's threshold and won't until afterwards. Aim to be comfortably clear of the region where those quoted figures cluster, with margin for a hard paper and a bad morning.

3. Where the real data lives

Three official sources, all free, all more reliable than any blog:

The GMC's "Recent pass rates for PLAB 1 and PLAB 2" page, in the PLAB section of gmc-uk.org. This is the authoritative published record of recent sitting results.

The GMC's "State of Medical Education and Practice" report, published annually, which includes PLAB data with breakdowns by factors including country of qualification and attempt number. The accompanying workforce report carries related data on how PLAB-route doctors progress into UK posts.

The GMC's report "Influences on pass rates in PLAB examinations", which examines what actually correlates with performance rather than just reporting headline percentages.

A transparency note: gmc-uk.org blocks automated access, so we could not load those pages to quote their current figures directly, and we're not going to reproduce numbers we couldn't verify at source. That's exactly the failure mode this post is about. Go to gmc-uk.org and read the current figures yourself — it takes two minutes and it's the only version that's actually current.

4. Why published figures legitimately differ

Not all the variation is sloppiness. Some of it is real, and understanding it lets you read any figure correctly.

Which cohort is being counted. "Pass rate" might mean all candidates at a sitting, first-attempt candidates only, or candidates on their third or fourth attempt. First-attempt rates are consistently higher than all-attempt rates, because the resitting population is by definition made up of people who didn't pass. A guide quoting one and a guide quoting the other will disagree by a wide margin while both being accurate.

Which sitting. PLAB 1 runs multiple times a year across UK and international centres. Rates vary between individual sittings. An annual average and a single sitting's figure are different quantities.

Before or after 8 August 2024. Every PLAB 1 paper from that date is built on the GMC's MLA content map, to the same standard as the UKMLA AKT — replacing PLAB's own narrower blueprint. That's a change of scope and emphasis, so pre- and post-alignment figures aren't measuring the same exam. Several sources report early post-alignment sittings sitting at the lower end of the historical range, which is what you'd expect during a transition. Note that any figure from 2023 or earlier describes a retired exam.

Country of qualification and other factors. The GMC's own analysis breaks performance down by several candidate characteristics, so an overall figure and a subgroup figure will differ.

PLAB 1 confused with PLAB 2. They're separate exams with separate rates, and some guides blur them.

Pass mark confused with pass rate. The proportion of marks you need and the proportion of candidates who pass are unrelated quantities. Both get quoted as "around 65%", which is a coincidence of arithmetic and a reliable source of confusion.

5. Why the pass rate doesn't predict your result

Here's the part worth internalising, whatever the current figure turns out to be.

Suppose the pass rate is 62%. That is a fact about a population. It is not the probability that you will pass, and treating it as one is a base-rate error.

That population contains candidates who prepared for two weeks and candidates who prepared for five months. It contains people on their fourth attempt and people on their first. It contains doctors who never sat a single timed paper and doctors who sat six. You are not a random draw from that distribution — you are a specific person with a specific baseline and a specific plan, and both of those are things you control.

The pass rate is useful for exactly two things:

Calibrating seriousness. Roughly a third to nearly half of candidates don't pass, depending on the figure and cohort. That is not a formality, and it isn't an exam you can walk into on general clinical competence.
Understanding that resitting is common. If you fail, you are in a large group, not an unusual one.

It is useless for the thing people actually want it for, which is reassurance.

6. What actually predicts it

Measurable things, all of which you can act on:

First-attempt accuracy on fresh questions. The critical qualifier is first-attempt and fresh. Accuracy on questions you've seen before measures recall of a question bank, not clinical reasoning, and it inflates steeply with re-attempts. This is the single most honest number in your preparation, and many candidates never look at it.

Performance on full-length timed papers. 180 questions in 3 hours. Consistent, comfortable clearance across several timed papers made of unseen questions is the closest thing to a real predictor that exists. One good score isn't a trend.

Whether you finish. Timing failure is common and entirely preventable, and it's invisible until you sit a full paper under the clock.

Coverage breadth. Sampling the whole content map, including ethics and law, safeguarding, capacity, prescribing safety, statistics and health promotion — the domains almost everyone under-revises because they're less interesting than clinical medicine.

UK-pathway accuracy specifically. For doctors who trained elsewhere, the dominant error pattern is right diagnosis, wrong NHS management. It's systematic rather than random, which makes it efficiently fixable once you're measuring it. NICE guidelines for the UKMLA.

Calibration. Of the questions you felt confident about, how many were right? Confident wrongness is the most expensive category in any error log, because it never gets revised.

What doesn't predict it: total questions attempted, hours logged, a platform's readiness score, or how many recall PDFs you've collected. Why recalls in particular mislead you.

For how to build a plan around these, see how to pass PLAB 1, and for the honest arithmetic on question volume, how many questions you need.

7. If you fail: the attempt rules

Know these before you plan, because they're stricter than most candidates realise.

You may sit PLAB 1 until you have four failed attempts. After a fourth fail you must apply for a fifth and final attempt, and the GMC requires at least 12 months of additional learning and development after that fourth fail before it will consider the application — you cannot apply immediately. After five failed attempts you are no longer eligible. The same structure applies to PLAB 2. Check the current rules on the GMC's own "additional attempt" pages; the policy is theirs to change.

This arithmetic should shape your preparation. Attempts are finite, and the fourth one is expensive in years as well as fees.

It gets tighter from 2027: the GMC is closing PLAB 1 in Dhaka, Alexandria, Accra and Chennai from February 2027 and reducing international sittings from four a year to three. Fewer sittings means a longer wait between attempts. The 2027 changes.

After a fail, in order:

Read your feedback before buying anything. The breakdown tells you where you actually lost marks. Most people skip this and buy a new question bank instead, which is spending money to avoid information.
Work out whether it was knowledge, timing or pathways. Three different problems with three different fixes. A new bank fixes none of them by itself.
Change one thing deliberately — usually mock frequency, or the sourcing of your explanations.
Book the realistic date, not the soonest one. With three international sittings a year, a rushed resit can cost you the next window too.

Our retake guide covers the diagnosis in detail.

8. FAQ

What is the PLAB 1 pass mark? There isn't a fixed one. The GMC sets it per paper using the Angoff method plus one standard error of measurement, so it varies with the difficulty of your specific paper. Any guide quoting a single figure is either simplifying or repeating an unsourced number.

What is the PLAB 1 pass rate? It varies by sitting and by cohort. Third-party guides quote figures across a wide range — from around 55% to around 75% — depending on whether they count all attempts or first attempts, and which sittings. Read the current figures on the GMC's own "Recent pass rates for PLAB 1 and PLAB 2" page rather than trusting any blog, ours included.

Is PLAB 1 marked on a curve? No. It's criterion-referenced. There's no quota of passes and you're not competing against other candidates.

Do I need 65%? That figure circulates widely but isn't a published threshold, and the pass mark for your paper is set for your paper. Aim comfortably above the region those quoted figures cluster in rather than at a specific line.

Is there negative marking? No. Never leave a question blank — a blank is a guaranteed zero and a guess isn't.

Did the pass rate drop after the MLA alignment? Several sources report early post-August-2024 sittings at the lower end of the historical range, which is what you'd expect from a broader content map during a transition. We couldn't verify current figures at source, so check the GMC page.

Why do different websites give different numbers? Different cohorts, different sittings, pre- versus post-2024 exam, PLAB 1 confused with PLAB 2, pass mark confused with pass rate — and a lot of unsourced copying. Section 4 breaks it down.

How do I know if I'm ready? Consistent, comfortable clearance on full-length timed papers built from fresh questions, finishing inside the time, with no weak areas left. Not a percentage from a readiness model.

How many times can I sit PLAB 1? Four failed attempts, then an application for a fifth and final attempt requiring at least 12 months of additional learning and development after the fourth fail. No attempts beyond the fifth.

Is the UKMLA AKT pass mark the same? Same principle — criterion-referenced and standard-set rather than a fixed percentage — though the UK route's paper structure differs. How the UKMLA pass mark works.

Replace the guesswork with a number that's actually about you. Sit the free 50-question mock — fresh, map-aligned questions, every answer referenced to NICE or the BNF, with a per-specialty first-attempt breakdown. That figure predicts your result far better than any published pass rate. Or ten questions with no account.

The summary: the pass mark is set per paper by the Angoff method plus one standard error of measurement, so there is no fixed threshold — and the contradictory numbers you've been finding are mostly unsourced figures copied between guides. The pass rate is a fact about a population that includes people who prepared for a fortnight, and it tells you almost nothing about yourself.

Stop looking for the threshold. Measure your first-attempt accuracy on fresh timed questions, get it comfortably clear, and let the standard-setters worry about where the line falls.