Although women suffering from advanced cancer of the breast or ovary are unlikely to be cured, several active agents are available that can prolong their lives. The use of these agents is based on demonstrated benefit in large randomized clinical trials, and the clinical activity of these chemotherapy regimens is initially high, with 60%-70% of patients responding. Unfortunately, their benefit in the second-line setting is often limited, with less than 30% of patients showing significant disease response. Thus some 70% of patients may undergo ineffective treatment during the course of their disease, while still suffering from significant chemotherapy-related toxicity. Having some foreknowledge of a given agent's expected result before its administration would therefore benefit the individual patient. In vitro drug response testing, first developed to assist in the selection of antibiotics for patients with bacterial infections, has recently been demonstrated to accurately predict how cancer patients will respond to chemotherapy. This review discusses the historical development of in vitro testing for cancer patients, some of the pitfalls encountered, and offers an assessment of their current utility. Results of various clinical trials that evaluated correlations between in vitro tumor response and clinical outcomes are described. These data suggest that in vitro drug response assays can accurately predict drug resistance and can identify patients who are more or less likely to benefit from a given agent.