Medico-economic evaluations estimate, for a given health technology, the added cost and the clinical benefit compared to a reference strategy. The objective here is to analyze the criteria used to measure clinical benefit as the basis for market access and reimbursement decisions for drugs in oncology both in France and in Europe. Prolonged overall survival is the criterion of choice to demonstrate the benefit of an anticancer drug; a survival gain of 2 to 3 months or more would be considered as relevant for a new product versus the comparator. In the absence of survival benefit or mature data on survival, progression-free survival or symptom-free survival and the availability of alternative curative treatments, decrease in drug toxicity and quality of life improvement may be considered. Differences in clinical benefit assessment between regulatory agencies and payers are not specific to France. Case studies show that it is difficult to find a consistency in reimbursement and pricing decisions and to identify factors that may fully explain reimbursement decisions when survival benefit is not demonstrated.