An important question in radiobiology is the relationship between primary DNA damage and chromosomal aberrations. What determines the chromosomal aberration frequency, especially in radiosensitive cells? Much evidence points to the double-strand break (dsb) as the critical lesion, however there is controversy over whether it is the initial induction, repair or residual dsb which determine of the level of expression of chromosome damage. The picture is further complicated by the fact that chromosome damage can be measured at several levels e.g. at metaphase, as micronuclei and as prematurely condensed chromosomes. Differential frequencies of chromosome damage are measured in different cell lines. Repair and residual dsb may play a role in metaphase aberrations when cells are exposed in G1, but in irradiated G2 cells the differential frequencies do not depend on repair of dsb or on the residual level of dsb since a difference in the cell lines is observed at short intervals after irradiation, and in radiosensitive cell lines where there is no deficiency in the repair of dsb, e.g. ataxia telangiectasia cells. Thus, at least in G2 cells, a mechanism involving 'conversion' of dsb into chromatid breaks is proposed. There are a number of possible reasons for high conversion of dsb into chromatid breaks including altered chromatin structure, high chromosome condensation rates and covalent closure of chromosome ends.