This paper explores usage of the concept of abnormality in medical genetics and proposes directives for more careful usage of this concept. The conceptual difficulties are first explored, then a model is developed to assess actual usage, followed by analysis of a sample of genetic textbooks and genetics literature. It appears that fact and valuation are often intermingled, that reference standards used to define 'genetic abnormalities' are often not clear and that the concept of abnormality is often used independent of the degree of certainty with which the altered genetype develops into a (seriously) harmful phenotype. On the basis of these findings it is argued that more restraint and more careful use of the concept of genetic abnormality of medical genetics is appropriate as well as more agreement on the use of reference standards.