OBJECTIVE The problem of dissimulation by applicants when self-report tests of personality are used for job selection has received considerable attention in non-medical contexts. Personality testing is not yet widely used in medical student selection, but this may change in the light of recent research demonstrating significant relationships between personality and performance in medical school. This study therefore aimed to assess the extent of self-enhancement in a sample of medical school applicants. METHODS A within-subjects design compared personality test scores collected in 2007 for 83 newly enrolled medical students with scores for the same students obtained on the same personality test administered during the selection process 4 months previously. Five factors of personality were measured using the International Personality Item Pool and mean differences in scores were assessed using paired t-tests. RESULTS At the time of selection, the personality scores of successful applicants were similar to those of candidates who were not accepted (n=271). Once selected, the medical students achieved significantly lower scores on four of the five personality factors (conscientiousness, extroversion, openness to experience, agreeableness) and higher scores on the fifth factor (neuroticism). Of the selected students, 62.7% appeared to have 'faked good' on at least one of the five factors measured. CONCLUSIONS Applicants to medical school are likely to dissimulate when completing self-report tests of personality used for selection. The authors review the evidence as to whether such dissimulation reduces construct and predictive validity and summarise methods used to reduce self-enhancement in applicant samples.