According to Leape & Berwick (2000) the need to reduce medical errors is ‘obvious and the mandate is clear’. My article questions this assertion. I go beyond the unknown incidence of medical errors in a general medical population to suggest that the meaning of medical errors is itself equivocal. I contest the assumption that the ‘wrongness’ of medical errors is always problematic, arguing instead for a distinction between desirable errors and undesirable errors. This distinction takes into account the consequences of errors, and why they may occur. Reasons include the inappropriateness of two cultural contexts – evidence-based medicine and continuous quality improvement – within which patient safety standards can be constructed and hence, medical errors can be defined.