Is UCAT Getting Harder? What the Evidence Actually Shows

UCAT is developed by Pearson Vue and the UCAT Consortium, which includes all UK medical schools that use it in admissions. The test development process includes extensive question piloting and psychometric analysis to ensure that different versions of the exam — sat on different days during the testing window — produce comparable score distributions. This process is called equating, and it is specifically designed to prevent one cohort of students from being systematically disadvantaged by a harder version of the test. This means that if the questions on a given day genuinely are more difficult, the scoring distribution adjusts to reflect this. A score of 700 in one sitting period is intended to reflect the same performance level as a score of 700 in another sitting period within the same year. The equating process is not perfect, but it is robust enough that UCAT scores in a given year are broadly comparable regardless of your specific sitting date.

While the difficulty calibration process means UCAT does not become harder in a raw sense, several things have changed over time that affect how demanding students find it. The applicant pool has become more prepared — as commercial revision resources have proliferated, average preparation quality and quantity has increased, meaning the effective competition for high scores has intensified even if the exam itself has not become harder. The 2026 format change is the most significant structural development in recent years. The removal of Abstract Reasoning and the expansion of Decision Making change the skills profile required to perform well. Students who were strong on AR and weaker on logical reasoning are now at a disadvantage relative to the previous format. Students with strong analytical reasoning and data interpretation skills are now relatively advantaged.

The most practical implication of the difficulty calibration context is this: do not compare your practice scores to historical UCAT score data from before 2026 without adjusting for the format change. A 700 average in the old five-subtest format is not the same as a 700 average in the new three-subtest format, both because the subtest composition has changed and because the score scale has changed. For 2026, use score resources and percentile guides that are explicitly calibrated to the 2700-maximum format. MediSpoon's score analysis tools are built for the 2026 format and give you percentile context based on projections from the new score distribution.