Background: Some authors have pointed out that setting up an exam with random questions can impair student performance. Since the COVID-19 pandemic and compulsory social isolation, postgraduate medical education activities became virtual and exams were implemented online. Objective: The aim of the study was to analyse whether the random order of questions has any effect on test results. Method. Two exams were written: one on the Prevention of Cardiovascular Disease and another on Heart Valve Diseases. For each topic, two questionnaires were designed: a version with the questions in logical order and according to the estimated difficulty and another with the same questions randomly distributed. Each exam had 50 multiple-choice questions with 3 options, and the maximum possible score was 50 The exams were taken on the Moodle platform, with synchronous modality, and the time available was 75 minutes. The results were expressed as scores obtained (range and central tendency) and according to the index of difficulty of the questions. Results. The number of respondents was 284 residents, students of the Biannual Cardiology Course: two 1st year groups (Prevention topic) and two 2nd year groups (Valve diseases). There was no difference between the results of the two versions of the same exam. Conclusions. This study lacks sufficient power to support either way of question order in clinical sciences exams, giving rise to new queries.