Breast self-examination (BSE), increasingly promoted as a good health habit, is gaining recognition by surgeons as a significant factor in tumour size and staging at diagnosis. Clinically-based reports leave unanswered important behavioural questions and pay scant attention to the quality of the teaching or of the practice of BSE. This paper argues that if BSE is to be used as the coarse-screening modality of which it seems to be capable, we must apply to it the same criteria as to other screening modalities. Since BSE ipso facto must be undertaken by the women herself, it is a screening in which subject behaviour is crucial. Promotion of BSE requires not only behaviour change but also maintenance and support. Behavioural aspects are discussed in relation to a recent major study of the teaching and practice of BSE.