Chinese hamster ovary cells were synchronized at the G(1)/S-phase boundary of the cell cycle and were pulse-labeled with (125)I-iododeoxyuridine 30 min after they entered the S phase. Cell samples were harvested and frozen for accumulation of (125)I decays during the first and second G(2) phase after labeling. Cell aliquots that had accumulated the desired number of decays were thawed and plated for evaluation micronucleus formation and cell death. Cells subjected to (125)I decays during the first G(2) phase after labeling exhibited single-hit kinetics of cell killing (n = 1, D(0) 41 decays/cell). In contrast, decays accumulated during the second G(2) phase killed cells with dual-hit kinetics (n = 1.9, D(0) 81 decays/cell). A similar divergence in the action of (125)I was noted for micronucleus formation. These findings indicate that the effects of (125)I varied depending on whether the decays occurred in daughter DNA (first G(2) phase) or parent DNA (second G(2) phase). Control studies with external X rays showed no such divergence of the action of radiation. To account for this paradox, a model is proposed that invokes higher-order chromatin structures as radiation targets. This model implies differential spatial arrangements for parent and daughter DNA in the genome, with DNA strands organized such that a single (125)I decay originating in daughter DNA damages two targets during the first G(2) phase, but identical decays occurring during the second G(2) phase damage only one of the targets.