The purpose of this paper is to clarify the (1) objectives of screening programs, particularly for cancer, (2) characteristics of diseases suitable for screening, (3) features of suitable screening tests, (4) levels of program evaluation, and (5) evidence on the advantages and disadvantages of screening. Diseases suitable for screening are those leading to serious morbidity and high mortality, those with a prolonged preclinical phase in their natural course, and those for which effective therapy is available following early diagnosis. Useful screening procedures have documented validity, acceptability, and safety and are performed with relative ease and modest cost. These generic issues in cancer screening are examined in the context of screening for cervical, breast, and lung cancer. Each of these cancers illustrates how to weigh different kinds of evidence when advocating or rejecting screening.