Best evidence medical education (BEME) is an attempt to examine systematically the quality of the evidence available to address educational issues, using six dimensions: quality, utility, extent, strength, target and setting. The present paper is a critical examination of some of the assumptions of this method. Five basic points are made: (1) educational research is not necessarily inferior to clinical research; (2) methodological rigour is not unidimensional; (3) educational interventions cannot be easily standardized; (4) attempts to examine the strength and extent of evidence presuppose a single world-view; (5) however, despite these concerns, generalization is possible for many questions.