BACKGROUND:The assessment of medical students' clinical skill acquisition is variable and is driven by differences in curricula and health services requirements. This project aimed to ascertain whether different assessment models impact on students' performance of as intravenous (i.v.) cannulation. Recency of practice was also explored as a factor influencing performance. METHODS:A total of 137 students in the first clinical year (years 2 or 3 of a 4-year course) of their medical degree, from four regional clinical schools, participated in a multiple-station mock objective structured clinical examination (MOSCE). Intravenous cannulation was one of the stations examined. Fifty-nine students came from a model that required the assessment of i.v. proficiency during their clinical year (model 1). Seventy-eight students came from a model that required no assessment of i.v. proficiency (model 2). Students reported their most recent clinical i.v. cannulation experience relative to the MOSCE. RESULTS:The MOSCE pass rate of 73% for students in model 1 was significantly higher than the corresponding MOSCE pass rate of 45% for students in model 2. There was a highly significant association between assessment model and MOSCE pass/fail rate. The assessment model was also highly associated with recency of practice. The assessment of medical students' clinical skill acquisition is variable CONCLUSIONS: The results support an assessment model that promotes the continuing clinical practice of i.v. cannulation. Integration of this model will require innovative approaches by staff and collaboration with affiliated organisations.