[The low doses of radiation: Towards a new reading of the risk assessment].

From Hiroshima bomb explosion data, the risk of radiation-induced cancer is significant from 100 mSv for a population considered as uniform and radioresistant. However, the recent radiobiological data bring some new elements that highlight some features that were not taken into account: the individual factor, the dose rate and the repeated dose effect. The objective evaluation of the cancer risk due to doses lower than 100 mSv is conditioned by high levels of measurability and statistical significance. However, it appears that methodological rigor is not systematically applied in all the papers. Furthermore, unclear communication in press often leads to some announcement effects, which does not improve the readability of the issue. This papers aims to better understand the complexity of the low-dose-specific phenomena as a whole, by confronting the recent biological data with epidemiological data.