BackgroundPatient-centred care is a model of care that demands healthcare providers change their focus from the disease to the patient and his or her perceived physical and psycho-social needs. This model requires healthcare workers to listen actively and to have effective communication skills and well-developed levels of empathy.AimThe aim of this study was to determine the suitability of the Jefferson Scale for Empathy (JSE-S) as a valid test for empathy in third-year medical students at a South African university and also to determine the baseline level of empathy in this same group of students.SettingThe study took place at a medical school in the Western Cape, South Africa. This medical degree (MB ChB) is a 6-year programme. Students are first exposed to patients within their second year of training, but it is during their third-year that they start their clinical rotations. We wanted to test whether our empathy training would give students the necessary skills and enable them to establish good empathic communication habits in order to prevent a fall in empathy during this vulnerable period.MethodsThis article explores the suitability of the student version of the JSE-S as a valid test for empathy, within the South African medical school context. We briefly discuss the psychometrics and the scores against what is already known in countries like ours, specifically, developing nations where cultural and language differences exist in the student populations. Furthermore, we explore whether the JSE-S is a valid scale for pre- and post-intervention measurement of medical student empathy within our context and discuss the limitations of self-assessment. We also report on baseline levels of empathy in third-year medical students.ResultsTwo hundred and six third-year medical students (69% females) completed the JSE-S prior to the intervention. Females and students aged 25 years and older had significantly higher scores than males and those 22 years old or less. The mean JSE was 109.98 (SD = 12.54), which is lower than most internationally reported scores. The Cronbach’s alpha coefficient was 0.81, indicating scale reliability and consistency, but graded item response testing highlighted variance in three reverse-scored questions.ConclusionThe JSE-S is an appropriate and valid scale for measuring levels of empathy in undergraduate medical students in South Africa. However, language may need to be clarified in the negatively phrased items.