Three critical aspects govern successful Phase II cancer chemoprevention trials--well-characterized agents, suitable cohorts, and reliable intermediate biomarkers for measuring efficacy. Requirements for the agent are experimental or epidemiological data showing chemopreventive efficacy, safety on chronic administration, and a mechanistic rationale for the chemopreventive activity observed. The cohort should be suitable for measuring the chemopreventive activity of the agent and the intermediate biomarkers chosen. Also, many cohorts proposed for Phase II trials are patients with previous cancers or premalignant lesions. For such patients, the trials should be conducted within the context of standard treatment to avoid unusual risks. The criteria for biomarkers are that they fit expected biological mechanisms (i.e., differential expression in normal and high-risk tissue, on or closely linked to the causal pathway for the cancer, modulated by chemopreventive agents, and short latency compared with cancer), may be assayed reliably and quantitatively, measured easily, and correlate to decreased cancer incidence. They must occur in sufficient incidence to allow their biological and statistical evaluation relevant to cancer. Since carcinogenesis is a multipath process, single biomarkers are difficult to validate as surrogate endpoints, as they may appear on only one or a few of the many possible causal pathways. Panels of biomarkers, particularly those representing the range of carcinogenesis pathways, may prove more useful as surrogate endpoints. It is important to avoid relying solely on biomarkers representing isolated events that may or may not be on the causal pathway or otherwise associated with carcinogenesis. These include markers of normal cellular processes that may be increased or expressed during carcinogenesis, but are nonspecific.(ABSTRACT TRUNCATED AT 250 WORDS)