LEK exam results: what score you actually need, and why English-language candidates score lowest
96.8% pass, so the 112-point threshold decides nothing — your ranking does. Five sessions of CEM data, estimated percentile bands, and the statistic nobody writes about: English-language candidates average 133.65 against a national 157.09.
kursnalek editorial team
Good luck to everyone sitting the LEK on 5 September. The rest of this article is for those planning the spring session and wondering what number to actually aim for.
Because “passing” and “scoring well enough to matter” are two different targets, and the gap between them is several dozen points.
Almost everyone passes
| Session | Sat | Pass rate | Mean | % | Highest score |
|---|---|---|---|---|---|
| spring 2026 | 11,030 | 96.8% | 157.09 | 78.5% | 191 |
| autumn 2025 | 9,850 | 98.3% | 164.11 | 82.1% | 197 |
| spring 2025 | 10,714 | 98.5% | 162.88 | 81.4% | 198 |
| autumn 2024 | 8,852 | 95.7% | 153.80 | 76.9% | 186 |
| spring 2024 | 9,818 | 97.9% | 163.47 | 81.7% | 194 |
Across five consecutive sessions the pass rate never fell below 95.7%. The 112-point threshold is a formality — more than nineteen candidates in twenty clear it.
At the same time the standard deviation sits at roughly 20 points. The exam barely filters anyone out, but it sorts everyone very precisely — and it is that ordering, not the pass, that enters the residency qualification procedure.
The statistic nobody writes about
If you took your degree in English, or abroad, the national averages do not describe you.
| Group | Sat | Mean | Pass rate |
|---|---|---|---|
| English-language sitting | 334 | 133.65 | 81.7% |
| Foreign universities | 911 | 141.92 | 89.6% |
| National average | 11,030 | 157.09 | 96.8% |
| Best-performing university | 800 | 163.34 | 99.0% |
Roughly one in five English-language candidates does not pass, against one in thirty nationally. The gap between the English-language cohort and the strongest Polish faculty is nearly thirty points — a full one and a half standard deviations.
This is not a comment on the candidates. It is a predictable consequence of preparing for a Polish state examination from outside the Polish teaching system: the published question base, the archive of past sessions and most preparation material exist in Polish first. Sitting in English means working from a thinner shelf.
It also means the gap is more closable than the numbers suggest. It is the same paper, translated.
There are more places than candidates. The scarcity is elsewhere
The spring 2026 intake offered 4,039 residency and 8,221 non-residency training places — over twelve thousand against roughly eleven thousand candidates sitting the LEK in a session. Nationally, places are not scarce.
They are scarce in a particular place, in a particular specialty. The pool is divided across sixteen voivodeships and several dozen fields, so the most sought-after specialties can come down to one or two places in a given region. At that point belonging to the top quarter means nothing — what counts is being first on the list among everyone who applied for that field in that location.
This resolves the apparent contradiction around “almost a hundred percent”. Both statements are true at once:
- To get a residency at all — passing and applying broadly is enough. An application allows up to fifteen ranked choices.
- To get that one specialty in that one city — you have to outrank everyone who wants the same thing. With one place and a dozen applicants, that means a score from the very top of the distribution.
So the number worth aiming for depends on how rigid your target is. The narrower your choice of field and location, the higher you need to score — and the less comfort there is in knowing the national mean is 157 points.
What each score is worth
The bands below are an estimate for spring 2026 (mean 157.09, standard deviation 19.57), assuming an approximately normal distribution. CEM does not publish percentiles — the link to a graphical score distribution on its statistics page points to a file that no longer exists — so treat these as an approximation, not a measurement.
| Score | Percentage | Approximate standing |
|---|---|---|
| 157 pts | 78.5% | middle of the field |
| 170 pts | 85% | top quarter |
| 180 pts | 90% | top ~12% (~1,330 candidates) |
| 190 pts | 95% | top ~5% (~510 candidates) |
| 196 pts | 98% | top ~2% — session-record territory |
The last row is worth reading carefully, because it takes pressure off. 98% means 196 points — the highest score in the country in spring 2026 was 191. The two-year record is 198 points, set in spring 2025.
Which means the “almost a hundred percent” figure that circulates in residency conversations sits above the national record. The bar that actually changes your standing is lower than people say. The entire top quarter falls between 170 and 190 points, and that is where the ordering is decided.
Thirteen questions
The distance from the mean of 157 to a score of 170 is thirteen questions. From 170 to 190, another twenty.
That is the whole stake — not hundreds of additional hours, but thirty-odd specific questions the average candidate gets wrong. This is why working through the material front to back is expensive: half the paper consists of questions almost everyone answers correctly, and revising those moves nobody up the ranking.
What a question that defeats the field looks like
The three hardest items of spring 2026, with the full response distribution published by CEM.
Question 128 — 7.7% correct
Infection with the rabies virus will not occur: A. from an infected bat by the airborne route — 73.2% B. from an infected bat via the conjunctiva or nasal mucosa — 1.1% C. through corneal transplantation from an infected donor — 17.0% D. haematogenously, as a result of viraemia — 6.6% (correct) E. through saliva contaminating broken skin — 2.2%
Three quarters of candidates ruled out the airborne route. Aerosol transmission from bats is in fact documented, while viraemia does not occur in rabies — the virus travels along nerves. The trap is a negation in the stem combined with a fact that sounds implausible and is true.
Question 14 — 10.5% correct
A 75-year-old woman with atrial fibrillation, on apixaban, underwent percutaneous angioplasty of the left anterior descending artery with stent implantation following myocardial infarction. Dual antiplatelet therapy with aspirin and clopidogrel was started. After one week you should: A. discontinue aspirin — 8.5% (correct) B. discontinue clopidogrel — 22.7% C. switch apixaban to warfarin — 10.6% D. continue therapy for 6 months — 25.3% E. continue therapy for 12 months — 32.8%
The two “continue” options together drew 58.1% — the instinct that longer is safer. The point, however, is to shorten triple therapy as quickly as possible in a patient at bleeding risk: aspirin goes first, leaving clopidogrel alongside the anticoagulant.
Question 7 — 15.9% correct
An ACE inhibitor is indicated in a patient with heart failure and a left ventricular ejection fraction of 35%. After starting the drug, renal excretory function should be checked after: A. 3–5 days — 4.8% B. 1–2 weeks — 14.3% (correct) C. 3–4 weeks — 54.2% D. 6–8 weeks — 9.0% E. no monitoring required — 17.7%
More than half the field pushed the check two weeks too far out. This is a question about vigilance rather than theory: deterioration in renal function after starting an ACE inhibitor shows up within the first days, so a test a month later detects it too late.
None of these sits in an exotic corner of the syllabus. Rabies, triple therapy, ACE inhibitors — routine material. What failed was precision, not breadth.
Who actually fails
| Group | Failed |
|---|---|
| Graduated more than 2 years ago | 8.2% |
| Sitting for the first time | 4.6% |
| Graduates overall | 4.2% |
| Still students | 2.6% |
| Resitting to improve | 2.2% |
Risk rises with distance from graduation, not with the number of attempts. Someone resitting to improve their score outperforms a first-time candidate who has been away from study for a few years.
Where to start
The complete spring 2026 session — the same 200 questions, the same time limit — is available here free of charge. You will see your score and a full review of every answer, so you can place yourself against the 157.09 national mean and check whether you catch the three questions above.
Take the free LEK mock exam — spring 2026 session →
Data source: Centrum Egzaminów Medycznych — aggregate LEK session statistics (cem.edu.pl/lep_s.php) and the question base with item-level psychometrics (cem.edu.pl/pytcem). The percentile bands are estimated from a normal distribution using the mean and standard deviation published by CEM; they are not published figures. CEM does not release percentiles, and the link to a graphical score distribution on its statistics page resolves to a missing file.
Frequently asked questions
What is a good LEK score?
The pass mark is 112 out of 200, but almost everyone clears it — 96.8% passed in spring 2026. What matters is your position in the residency ranking. As a rough guide for that session: 170 points (85%) puts you in the top quarter, 180 points (90%) in roughly the top 12%, and 190 points (95%) in the top 5%.
How do English-language candidates perform on the LEK?
They are the lowest-scoring group in the country. In spring 2026, 334 candidates sat the exam in English, averaging 133.65 points with a pass rate of 81.7% — against a national average of 157.09 and a 96.8% pass rate. Roughly one in five did not pass, compared with one in thirty nationally.
Is 98% achievable on the LEK?
98% means 196 of 200 points — record territory. The highest score in the country in spring 2026 was 191 (95.5%), and the two-year record is 198 points, set in spring 2025. The useful news is that the bar which actually changes your standing sits lower: the top quarter of the field begins around 170 points, and the top 5% at 190.
Which LEK session was the hardest recently?
Of the last five sessions, autumn 2024 had the lowest mean at 153.80 with a 95.7% pass rate. Spring 2026 was the second hardest: a mean of 157.09 against 162–164 points in the three preceding sessions.
Does the LEK score affect residency placement?
Yes, directly. Your result is converted into a percentage of the 200-point maximum and that percentage feeds the ranking list used in the qualification procedure. There is no statutory cut-off per specialty — placement depends on how many candidates apply for the same specialty in the same voivodeship.
Sit a mock exam from the CEM base
A free test built from official CEM questions — 200 from the Spring 2026 session, with your score and the correct answer after each one. The exam itself is in Polish.