Summary This retrospective analysis of psychological predictors of attendance studied the women from the annual screening arm of the United Kingdom Coordinating Committee on Cancer Research (UKCCCR) trial of annual screening mammography for the early detection of breast cancer. Some women attended screening at the first invitation in year 1 (attenders), others did not attend for screening at any time (non-attenders), whereas a third group delayed attending until year 2 (ambivalent attenders). A total of 147 women were recruited to the study: 80 attenders, 28 non-attenders and 39 ambivalent attenders. It proved extremely difficult to contact non-attenders to take part in the study. Non-attenders were significantly more depressed on the Hospital Anxiety and Depression Scale; had experienced more miscarriages, stillbirths or terminations of pregnancy; were less knowledgeable about mammography; and were displeased to have received an invitation to screening. Whereas non-attenders are unlikely ever to attend breast screening because of their long-standing attitudes and preferred coping styles, ambivalent attenders may become more amenable to screening with the passage of time. In this study such women were persuaded to attend in year 2 with a simple, cost-effective intervention: an additional invitation letter after a year.Keywords: attendance; breast screening; depression; perinatal losses; psychology Low take-up rates for breast screening examinations pose a serious challenge for health promotion. Attenders and non-attenders for mammography have now been compared in a variety of UK, European and US settings, and the studies of Fallowfield, Rimer, Vernon, Frazier, Sutton and Polednak are useful exemplars.