Assessment of competence has made considerable progression in medical education in the last few decades. In essence these developments were driven by the desire to make assessment more authentic, i.e. to strive for a closer match with the complex reality of professional medicine. Many lessons have been learned from research on assessment. This presentation reviews these lessons by discussing issues around reliability (is our measure precise?), validity (is our measure accurate) and educational consequences (how does assessment impact on learners and teachers?). It is argued that the way forward for assessment is towards educational design strategy. This can be achieved by less emphasis on a psychometric approach where measurement characteristics of individual instruments determine assessment programmes. Instead, the design of assessment programmes should be driven by the desired effects of the programme on the education system. To support this presentation a portfolio of papers with overviews is available that allows the interested person to explore some assessment strategies further.