Resistance to multiple chemotherapeutic agents is a common clinical problem in the treatment of cancer: such resistance may occur in primary therapy or be acquired during treatment. The most commonly used antineoplastic agents in the treatment of disseminated breast cancer are adriamycin, methotrexate and cyclophosphamide. Cell lines selected for resistance to adriamycin often develop cross-resistance to structurally dissimilar antineoplastic drugs with different mechanisms of cytotoxic action; this phenomenon has been called pleiotropic or multidrug resistance (MDR). In vitro models of MDR have shown that this type of resistance is accompanied by a decrease in cellular drug accumulation, mediated by the over-expression of a 170 kD plasma membrane glycoprotein referred to as P170. Glycoprotein P170 is an energy-dependent multidrug efflux pump, whose activity can be inhibited in vitro by a variety of agents including verapamil, quinidine and reserpine. P170 is over-expressed also in some human malignancies, and evidence exists about its role in examples of clinical resistance in vitro. Clinical trials using verapamil, a calcium channel blocker which selectively enhances drug cytotoxicity in MDR cell lines, have been prompted for leukemia and ovarian cancer. In addition other approaches are the subject of current preclinical investigations. Several observations as well the phenomenon of "atypical" MDR in cell lines which do not overexpress P170, suggest that also other factors are involved in multidrug resistance. Qualitative or quantitative changes in the activity of topoisomerases, protein kinase-related systems and glutathione S-transferase, may confer pleiotropic resistance. As the role of these genes and their regulation is clarified, they may also serve as useful targets for pharmacologic intervention in the treatment of drug-resistant human tumors. The mechanisms involved in resistance to methotrexate and cyclophosphamide are less studied, particularly in vivo samples. Methotrexate resistance is probably a complex multifactorial phenomenon; in some cases it is due to an increase in the expression of the drug target dihydrofolate reductase, often as a result of gene amplification, but in other cases a transport defect of the methotrexate or alterations of the activity of different enzymes have been reported. Cyclophosphamide (CP) resistance has been attributed to an increased activity of two different enzymes, glutathione S-transferase, also involved in MDR phenotype, and aldehyde dehydrogenase, which catalyzes inactivation of CP in non cytotoxic metabolites. This paper reviews the current state of our knowledge of chemo-resistance and the utility of available markers to identify potentially resistant tumors in vivo; the strategies that might be used to overcome this phenomenon are also described.