The ultimate value of a screening program based on the response of cell lines in vitro will depend on the demonstration of a strong correlation between in vitro and in vivo responses to cytotoxic drugs. However, marked discrepancies in the predictive value of in vitro chemosensitivity assays have been described, suggesting that factors other than the inherent chemosensitivity of tumor cells significantly influence the outcome of chemotherapy in vivo. These factors include the influence of drug pharmacokinetics and metabolism, together with numerous problems associated with the biology of solid tumors in vivo (e.g., drug penetration barriers, proliferation gradients, and microenvironmental conditions). These additional factors may be highly significant in explaining the site-dependent nature of the responses of solid tumors to cytotoxic drugs, the poor prediction of responses in experimental tumor models, and the differences in the responses of multicellular spheroids and monolayers. There discrepancies suggest that the selection of compounds for phase II clinical trials on the basis of disease-specific activity in vitro may be premature. [J Natl Cancer Inst 82: 1457-1468, 1990].