Many studies have shown that screening for breast cancer can reduce mortality from the disease. Mammography has come to be regarded as the screening method of choice, but evidence suggests that physical examination (PE) is at least as effective in reducing mortality. Mammography detects many non-infiltrating and small, non-palpable tumours, but we do not know whether these would ever cause symptoms or threaten the woman's life. It is doubtful whether the time gained by early mammographic detection confers any survival benefit over PE detection. PE has substantial advantages over mammography in terms of human and economic costs. The question we should be asking is not how to refine mammographic screening but whether we need it at all.