The goal of cancer screening is early detection and treatment of disease with a consequent reduction in mortality. The difficult task of evaluating whether a particular screening program can achieve this goal is the theme of this article. Our focus is on the two principal components of the screening process. The first component is the ability of the screening test to detect cancer early, while minimizing the number of false positive results. In this regard, the specificity of the test ordinarily must be very high, approaching 99%. The second component in the screening process is evaluating the overall impact of a screening program. For this, various study designs have been proposed. The best procedure, the randomized controlled trial, is discussed in detail, but alternative designs are also considered. The endpoint in any such study should be cancer-specific mortality. The problems inherent in other outcome variables, such as stage shift and increased length of survival due to lead time and length biases are addressed. Also considered are major difficulties in study design and evaluation, such as inadequate knowledge of the natural history of the disease, particularly of preclinical lesions. Key data items that should be collected for evaluation are listed.