The induction of DNA double-strand breaks (DSBs) by genotoxic treatment leads to high toxicity and genetic instability. Various approaches have been undertaken to quantify the number of breaks and to follow the kinetic of DSB repair. Recently, the phosphorylation of the variant histone H2AX (named gammaH2AX), quantified by specific immunodetection approaches, has provided a valuable and highly sensitive method to monitor DSBs formation. Although it is admitted that the number of gammaH2AX foci reflected that of DSBs, contradictory reports leave open the question of a link between the disappearance of gammaH2AX signal and DSBs repair. We determined gammaH2AX expression (i) in cells either proficient or not in DSBs repair capacity, (ii) after exposure to ionizing radiation (IR) or calicheamicin gamma1, a radiomimetic compound, (iii) and by three different immunodetection methods, foci numbering, flow cytometry or Western blotting. We showed here that gammaH2AX loss correlates with DSB repair activity only at low cytotoxic doses, when less than 100-150 DSBs breaks per genome are produced, independently of the method used. In addition, in DNA repair proficient cells, the early decrease in the number and intensity of gammaH2AX foci observed after a 2 Gy exposure was not associated with a significant change in the global gammaH2AX level as determined by Western blotting or flow cytometry. These results suggest that the dephosphorylation step of gammaH2AX may be limiting and that the loss of foci is mediated not only by gammaH2AX dephosphorylation but also through its redistribution towards the chromatin.