G. MATHE and R. K. OLDHAM'~ Institut de Cancerologie et d'Immunogem!tique, Hopital Paul Brousse, Villejuif Since the last war, cancer chemotherapy has been the object of very intensive and expensive research. Nevertheless, its development has been very slow, and its ultimate potential is today somewhat in doubt. In doubt because it does not cure any cancer patients except a) females carrying placental choriocarcinoma, a semi-allogenic tumor, in which case, cure may be in­ fluenced by immune rejection, and b) children suffering from Burkitt's tumor, where the probable reason for the cure is that all the neoplastic cells are in the cycle, which is a unique condition among all the human tumor varieties. Whether the long term survivors in acute leukemia, lymphomas, certain sarcomas and certain testicular tumors are "cures" will require longer follow-up. The idea that chemotherapy does not cure most cancer patients because all their neoplastic cells are not in cycle has led to the use of drug combinations. Whatever they are, "cocktail combinations" which are made up of drugs given according to any timing, or scientific combinations, based on pharmacodynamics, pharmaco­ kinetics or cell kinetics data, are more toxic than single drugs, and are all the more toxic as the number of drugs in the combination is increased.