OBJECTIVES: Breast cancer (BC) incidence has steadily increased in France over the period 1980-2005 by a mean +2,4 % yearly, whereas mortality decreased since the early 90’s by  a mean 1.3 % yearly. Two major factors may explain this apparent discrepancy: on one hand, the extension of breast cancer screening coverage has permitted an earlier diagnosis of cancer and, on the other hand, therapeutic progress has led to improved survival. Our objective was to quantify the relative role of these two factors in France over the period 1994-2005. METHODS: A simulation model was developed to extrapolate the evolution of the distribution of BC by stages at diagnosis as observed in the French registries in 1994 and 2006. The survival according to the stage as observed in 1994 was then adjusted through an exponential model in order to fit the observed breast cancer specific mortality rate as reported in the French vital statistics until the year 2005. RESULTS: In 2005, the absolute number of deaths for BC was 11,381 in France. Our model estimated the corresponding values that would be obtained respectively in absence of any modification of the stage distribution between 1994 and 2006 (12,698 in 2005 i.e. 11% increase), or in absence of any modification of survival by stage at diagnosis from 1994 onward (12,752 in 2005, i.e. 11.5% increase). The contribution of therapeutic progress in the downward trend of BC specific mortality is in the same order of magnitude as the screening coverage extension. These two factors acted in an independent and additive manner on mortality reduction. CONCLUSIONS: The contrasted evolution of BC incidence and mortality cannot be explained solely by the extension of screening. Therapeutic innovation played an important and equivalent role in the trends over the period 1994-2005.