Résumé : Two major mechanisms responsible for chromosome segregation errors are non-disjunction and chromosome loss, both leading to aneuploidy. Previous studies in our laboratory showed the existence of thresholds for the induction of chromosome non-disjunction and chromosome loss and the induction of apoptosis by microtubule inhibitors. From a mechanistic point of view one can expect that apoptosis contributes to the elimination of cells with premutagenic/mutagenic lesions. If aneuploid cells were eliminated by the induction of apoptosis below the threshold concentrations for chromosome loss and non-disjunction, the defined thresholds would not be applicable to cells unable to undergo apoptosis. The aim of this study was to investigate whether apoptosis was induced directly or indirectly as a response to aberrant chromosome segregation below the thresholds for the induction of chromosome loss and non-disjunction, as previously defined by us. Therefore, human lymphocytes were exposed in vitro to five concentrations of nocodazole and five concentrations of carbendazim representing the threshold concentrations for chromosome non-disjunction and chromosome loss, two concentrations below the lowest threshold and one concentration between the two threshold values. After 48 h exposure to the aneugens, induction of apoptosis was analysed by the annexin-V test. The frequencies of chromosome non-disjunction and chromosome loss were estimated in cytokinesis-blocked human lymphocytes in combination with FISH; this methodology was applied to whole cell cultures as well as to apoptotic and viable cell fractions obtained using magnetic annexin microbead cell sorting. Our results suggest that elimination of aneuploid cells does occur. However, the efficiency of disappearance of micronucleated cells is higher than for cells presenting chromosome non-disjunction. The correlation found between early apoptotic events and micronucleus formation could account, at least in part, for the specific elimination of aneuploid cells.