Reports from the Medical Licensing Assessment (MLA) Applied Knowledge Test (AKT) Board outline the delivery and performance of the AKT across UK medical schools. Introduced from the 2024/25 academic year, the MLA is a national requirement for medical graduates before joining the medical register.
Published annually, these reports summarise key activities, governance and outcomes, supported by contributions from medical schools and the General Medical Council. Frequently asked questions are provided to help interpret the findings and their implications.
Questions are written by trained academics from across UK medical schools. They go through multiple rounds of review for accuracy, clarity and fairness. The Item Development Group (IDG) oversees the growth and quality of the item bank, while the Exam Construction Group (ECG) oversees the construction of the exams according to a national sampling grid that is aligned to the GMC content map.
Refer to pages 15–17.
Key governance groups include:
Refer to pages 6–7.
Exams are constructed to reflect the knowledge, skills, and clinical areas defined by the GMC Content Map. While the Content Map list is not exhaustive, it identifies core conditions. This ensures that all exams fairly assess the knowledge required to safely enter the UK Foundation Programme.
Refer to pages 18–19.
The ECG uses a standardised sampling grid and automated construction algorithm to create the exams which are then reviewed manually to confirm alignment. This ensures all exams cover the same topics in similar proportions, maintaining fairness and consistency.
Refer to pages 18–19.
Each item undergoes multiple layers of peer review before it is available for use in exams and is then reviewed using psychometric data post exam. Items can be flagged, revised or retired if necessary.
Refer to pages 15–17.
Psychometric data help check the reliability, difficulty and discrimination of items. Analyses using Classical Test Theory and Item Response Theory inform decisions about pass marks and any potential item revisions.
Refer to pages 20–21.
Psychometric data help check the reliability, difficulty and discrimination of items. Analyses using Classical Test Theory and Item Response Theory inform decisions about pass marks and any potential item revisions.
Refer to pages 20–21.
The Standard Setting Group uses a Modified Angoff method, a criterion-referenced approach, where expert panels estimate the performance of a minimally competent candidate. Pass marks are reviewed for consistency across exams.
Refer to pages 20–21.
Students have access to a 200-item practice exam (blueprinted to the Content Map) and shorter practice papers. These materials help students become familiar with question styles and the online platform.
Refer to pages 9–11.
Medical schools, guided by a national policy framework, handle requests for reasonable adjustments (e.g., extra time, assistive technology). Decisions are made on a case-by-case basis, in line with equalities legislation.
Refer to pages 12–14.
Yes, individual medical schools apply their own mitigating circumstances policy to the exam guided by the AKT Policy Framework.
Refer to pages 12–14.
The AKT is centrally constructed and developed, with medical schools responsible for the local delivery of the AKT. Whilst there is a national policy framework, schools manage the delivery within the context of their university regulations and so each school is responsible for providing appropriate information to their students.
Refer to pages 9–11 and 12–14.
Cronbach’s alpha measures internal consistency. All AKT “main sit” exams had alpha values above 0.8, showing the exams are highly reliable and measure knowledge consistently.
Refer to pages and 21–22.
The AKT is designed as a minimum competence standard exam, so most students are expected to pass if they meet required knowledge levels. The high pass rate reflects that the exam reliably identifies those meeting this threshold.
Refer to pages and 21–22.
Results are highly consistent. Standard-setting panels show strong alignment with previous pilot data and performance across different schools and exam dates is comparable.
Small differences between schools are expected and reflect normal variation and differences in student cohorts. Everyone who passes the AKT meets the same national standard for safe medical practice.
Refer to pages 20–24.
Expert panels independently judge each item against a minimally competent candidate. This method reduces bias and ensures a national standard.
Refer to pages 20–21.
No items were removed in 2024–25 following post-exam review. Feedback on individual items is shared with the IDG to help inform the way those items are written in the future.
Refer to page 20.
Performance is very similar, with pass rates of 96.4% in penultimate-year sittings and 98.3% in final-year sittings, showing consistent standards across cohorts.
Refer to pages 22–24.
Small gaps were observed between groups sharing protected characteristics. For example:
The gaps are consistent with longstanding patterns seen in undergraduate and postgraduate medical assessments. The AKT does not appear to create new structural disadvantages but reveals existing patterns of disadvantage and provides a benchmark for schools to monitor and address awarding gaps.
Refer to pages 25–34.
The demographic data collected as part of the running of the AKT can allow analysis of how multiple characteristics (e.g. ethnicity and socio-economic background) may combine to affect outcomes, helping to identify and address compounded disadvantage. This analysis will not be part of the annual reporting of the AKT but will be carried out and published by research teams in due course.
Refer to page 41.
Schools receive benchmarking data to monitor awarding gaps and take targeted action. MSC supports best practice in assessment design and EDI across schools to reduce disadvantage. The MSC EDI Alliance, which includes membership from across every UK medical school, works to improve inclusivity and address inequity in medical education.
Refer to pages 25–30.
Schools can compare their students’ performance with national trends, track awarding gaps over time and develop strategies to support underperforming groups.
Refer to page 41.
The Rapid Response Team monitors live exams, provides immediate advice to schools and the MSC team, and ensures timely resolution of issues on exam days. Any problems are logged in real-time, decisions are communicated to schools, and feedback is used to improve future processes.
Refer to pages 23–24.
Automated exam construction, manual review and psychometric oversight ensure that all exam sittings are equivalent in content, coverage and difficulty.
Refer to pages 18–21.
Post-exam analysis, including comments on item performance, is shared with the IDG to improve the question bank and inform future exams.
Refer to page 20.
The AKT demographic survey collects novel data such as language proficiency and socio-economic background, enabling research into factors influencing performance.
Refer to pages 28–32 and 41.
By analysing combinations of characteristics (e.g. ethnicity + socio-economic status), researchers and schools can identify groups at higher risk of disadvantage and develop targeted support.
Refer to page 41.
Annual reports highlight trends, award gaps, and operational insights, guiding updates to exam content, policies, and student support to ensure fairness and reliability.
Refer to pages 25–30 and 41.
No. The purpose of the MLA is to ensure consistency, fairness and safety for patients, not to rank universities or their students. AKT and MLA results are a benchmark for competence, not a measure of institutional performance, and should not be used to judge or compare medical schools.
The AKT measures whether each student meets a national threshold for safe practice, not the overall quality of teaching at a medical school. Slight differences in pass rates between schools reflect normal variation in student cohorts, course timing and local assessment methods, not “better” or “worse” schools. The exam is pass/fail only, so using results to create national league tables would be misleading and uninformative.