Many human tumors are thought to develop along a multistep pathway in tissues that have encountered long periods of carcinogen exposure and thus have accumulated genetic hits in functional targets relevant to tumor evolution. The cumulative degree of genetic change is dependent on both exogenous (e.g., degree of carcinogen exposure) and endogenous factors (e.g., metabolism of procarcinogens, repair or misrepair capacity, proliferation properties of the tissue, capability of damaged cells to survive). Thus one approach to risk estimation is to measure the accumulated amount of genetic damage in a target tissue at risk for tumor development. Since one cannot predict the exact site of the future tumor, the risk assay must detect a generalized ongoing process of genetic instability from small, random biopsies. The technique of chromosome in situ hybridization involves the use of chromosome- or region-specific probes and provides an ability to directly visualize genetic change (e.g., random or clonal chromosome polysomy and monosomy) on thin tissue sections (where tissue architecture is maintained) or exfoliated cells. Analyses of normal and premalignant lesions adjacent to tumors (e.g., head and neck, lung, bladder, cervix, breast) have demonstrated that chromosome instability can be detected in the field of the tumor (i.e., in normal and premalignant cells in a tissue at 100% risk of tumor development) and the degree of chromosome instability increases with the degree of histologic progression toward cancer. Analyses of premalignant lesions (e.g., oral leukoplakia and erythroplakia from individuals at risk for aerodigestive tract cancer) by chromosome in situ hybridization have uncovered varying degrees of chromosome instability. However, approximately half of those individuals who showed a high degree of chromosome instability in biopsies subsequently developed aerodigestive tract cancer. Of interest, half of these tumors have developed away from the biopsied site, suggesting that the detection of a chromosome instability process in one aspect of the tissue might yield risk information for the total tissue field. These studies also suggest that chromosome in situ hybridization might be useful for identifying individuals with high tumor risk who might benefit from chemopreventive intervention.