Background Many medical schools have implemented curricula to teach non-technical skills, a personal set of complex social and cognitive skills which are grounded in human factors safety industries in and out of health. Consensus on how to assess these skills is lacking. This systematic review aimed to evaluate the evidence regarding non-technical skills assessments in undergraduate medical education, to describe the tools used, learning outcomes and the validity, reliability and psychometrics of the instruments. Given the discrete context, a focussed review model is being deployed. Methods Studies describing assessment methods as either the focus of the study or having non-technical skills assessment as an outcome measure of the research were considered. A standardized search of online databases was conducted and consensus reached on included studies. Data extraction, quality assessment and content analysis were conducted per Best Evidence in Medical Education guidelines. Results Nine papers met the inclusion criteria. Assessment methods broadly fell into three categories: simulated clinical scenarios, objective structured clinical examinations, and questionnaires or written assessments. Details of methodology were synthesised to support readers developing their own materials. Tools to assess non-technical skills were often developed locally, in 4 response to specific educational interventions, without reference to conceptual frameworks. Consequently, the tools were rarely validated, limiting dissemination and replication. The majority of studies achieved outcomes modifying knowledge and skills of participants. Two studies resulted in behavioural change and one resulted in change in practice. Conclusions There were clear themes in content and broad categories in methods of assessments employed, with the OSCE identified as most able to assess multiple related skills at once. The quality of this evidence was poor due to lack of theoretical underpinning, with most assessments not part of normal process, but rather produced as a specific outcome measure for a teaching based study. Data on validity, reliability and learning outcomes was not available so these questions cannot be addressed at this time. Whilst the current literature forms a good starting position for educators developing materials, there is a need for future work to address these weaknesses as such tools are required across health education.