The present study deals with retrospective estimation of radiation doses, among Estonian Chernobyl clean-up workers, by means of scoring stable translocations using the FISH technique. All persons investigated in this study were sent to the area in 1986 and the majority of them were selected to be among those with the presumably highest exposure doses. In spite of the selection the estimated average dose is between 0.2-0.3 Gy, thus not clearly above the officially permitted dose level of 0.25 Gy. Due to unforseen conditions during transport of the blood samples, both the number of persons available for analysis and the number of metaphases available for scoring were substantially reduced. However, unless this selection is linked with the potential aberration frequency of the cells involved, no bias is expected.