In cancer screening, it is considered mandatory not only to prove the efficacy of the screening test, but also, as a permanent commitment, to demonstrate and quantify the effectiveness of service screening in terms of mortality reduction when the test becomes part of a screening programme. In Germany, a nationwide organized and quality-assured mammography screening programme among 50-69-year-old women, a target population of about 10 million women, is currently implemented. The envisaged short implementation period of about 2 years raises the issue of how to estimate the effectiveness of the programme in terms of long-term mortality reduction. On the basis of sample size calculations, a strategy for evaluation of mortality reduction is outlined. Our approach compares breast cancer mortality in different geographical areas or temporally within geographical areas. The latter design appears useful also for established programmes to examine site-specific mortality reduction by screening in late-starting areas, which are frequently used as control areas of a geographical comparison, but rarely a subject of evaluation on their own. Emphasis has to be put on the collection of the required data from the very beginning of the programme. For this, performance of cancer registration and proper linkage with the screening programme must be enhanced. In the trade-off between data protection and public interest in quality-assurance of medical care, emphasis should be put on assurance of high-level medical care whose relevance may be counted in terms of the numbers of deaths prevented. It is from this the individual truly benefits.