Screening is a way of detecting disease early in an asymptomatic population. For cancer screening to be effective, there not only must be a test that will detect cancer earlier, but there also must be a treatment that will result in an improved outcome. The strongest evidence for screening benefit comes from randomized prospective trials with a decrease in mortality as the outcome. For women older than 69 years of age, there is no direct scientific evidence that screening mammography will decrease their mortality from breast cancer. If there is no direct evidence (positive or negative), what can we say about any potential benefit for older women? The Forum on Breast Cancer Screening in older women (held in Sturbridge, Massachusetts, in 1990 and sponsored by the National Cancer Institute and the National Institute of Aging) systematically reviewed a number of issues that were considered to have an indirect but positive impact on the benefit of screening mammography--incidence (which rises dramatically with age), mortality (greater in women older than 65), mammography detection (enhanced in breasts of older women), and elderly survival rates (the average women older than 65 lives long enough to benefit from screening). Unresolved issues were the proper interval for screening (12 vs. 24 vs. 33 months) and the extent to which clinical breast examination contributes to a decrease in mortality. Clinical research in the form of a national trial is needed, because the recommendations to initiate or continue screening mammography in women older than 65 is based not on scientific evidence but on opinion.