There are several reasons for studying the experimental basis of the combination of chemotherapy and radiation in lung cancer: Firstly, there is a proven small benefit for thoracic radiation on survival in limited stage SCLC treated with chemotherapy, and a possible benefit for chemotherapy in limited, inoperable NSCLC treated with radiotherapy. Secondly, the mechanisms of cellular resistance to radiotherapy and chemotherapy may differ and this may allow an increase in local control which itself may have an impact on survival and quality of life. The local control and toxicity will depend on the relative sensitivities of the tumor and normal tissues. There are many mechanisms whereby lung cancer and normal tissues are resistant to chemotherapy and radiation-induced damage. A decrease in cellular topoisomerase I or II has been associated with a decrease in cellular sensitivity to topoisomerase inhibitors. Mutations or deletions in the topoisomerase I or II gene are associated with resistance to topoisomerase I or II inhibitors, respectively. It seems probable that, in lung cancer, the major mechanisms of resistance to anthracyclines will not involve the MDR protein. Amplification of the dihydrofolate reductase gene is associated with methotrexate resistance in SCLC lines. The mechanism of cisplatin resistance is multifactorial although decrease of uptake is observed in the majority of cisplatin-resistant cells. Increase in cellular glutathione (GSH) and in GSH-S-transferases may be responsible for resistance to alkylating drugs. Increase in metallothianine 2a is associated with resistance to cisplatin in SCLC lines. These mechanisms may or may not also convey resistance to radiation. Analysis of in vitro chemosensitivity testing suggests that these tests have negative predictive value but are less satisfactory for positive prediction. At present, the methods are not recommended for routine use. The mechanisms of cellular resistance to irradiation are less well understood. Damage to DNA, especially double strand breaks, appears to be the lethal lesion. Enzymes involved in the repair of this lesion are not well characterized, nor is the biological basis of the differing radiation sensitivity of different tissues. There is