Introduction: Consideration has been given to the use of tests of moral reasoning in the selection procedure for medical students. We argue that moral orientation, rather than moral reasoning, might be more efficacious in minimising the likelihood of inappropriate ethical behaviour in medicine. A conceptualisation and measure of moral orientation are presented, together with findings from 11 samples of medical school applicants and students. Aim: To provide empirical evidence for the reliability and validity of a measure of moral orientation and to explore gender, age, cultural and educational influences on moral orientation. Methods: A questionnaire designed to measure a libertarian−dual− communitarian dimension of moral orientation was completed by 7864 medical school applicants and students in Australia, Israel, Fiji, New Zealand, Scotland and England and by 84 Australian psychology students between 1997 and 2001. Results: Older respondents produced marginally higher (more communitarian) moral orientation scores, as did women compared to men. Minor but significant (P < 0.05) cultural differences were found. The Israeli samples produced higher mean moral orientation scores, while the Australian psychology student sample produced a lower (more libertarian) mean score relative to all other samples. No significant change in moral orientation score was observed after 1 year in a sample of Australian medical school students (n = 59), although some differences observed between 5 cohorts of Australian medical students (Years 1–5; n = 234) did reach significance. Moral orientation scores were found to be significantly correlated with a number of personality measures, providing evidence of construct validity. In all samples moral orientation significantly predicted the moral decisions made in response to the hypothetical dilemmas embedded in the measurement instrument. Discussion: The results provide support for the conceptualisation of a libertarian−dual−communitarian dimension of moral orientation and demonstrate the psychometric properties of the measurement instrument. A number of questions concerning the use of such tests in selection procedures are considered.