As one of several explanations for class differentials in health, health selection has received remarkably little systematic attention in the inequalities debate. It is widely regarded as having (at best) a very minor role in the production of inequalities, and a theoretical debt to social Darwinism. This paper examines the validity of those assumptions in terms of the evidence which has emerged since the publication of the 'Black Report'. It is suggested that it is too easy to write off health selection as of little or no significance, and that reconceptualising the issue within a specifically sociological perspective owing much to labelling theory offers much greater potential for understanding the processes involved. From this perspective, health selection has many of the features of discrimination of the sort that characterises race and sex.