Fluorescent DNA probes are used to characterise the chromosome constitution of preimplantation embryos. FISH is used to select normal or balanced embryos in carriers of balanced chromosomal rearrangements, for embryo sexing or for aneuploidy screening in women of advanced age, who have had recurrent abortions or IVF failures. In most cases, FISH is performed on interphase blastomeres which are asynchronously dividing cells, that can be in G1, S or G2. However, a correct interpretation of a double FISH signal, which may correspond to a split signal, to a replicated chromosome region or to the presence of an extra chromosome is essential to establish an accurate diagnosis. To determine if the cell stage could influence the interpretation of FISH results, we compared the signal characteristics of one locus-specific probe, two different subtelomere region probes, and a centromere region probe in non-dividing Sertoli cells and in proliferating lymphocytes. Most cells had two signals per chromosome pair (i.e., a situation corresponding to G0 in Sertoli cells and to G1 or to a prereplication stage in lymphocytes). Nevertheless, in proliferating cells the percentage of nuclei with a number of signals different from the expected (two unreplicated chromosomes per pair) was different from that found in non-dividing cells (P < 0.05). It was estimated that 10.8% of double dots in dividing cells resulted from DNA replication. The sequence of replication was first the locus-specific region, second a telomere region, and third the centromere. In conclusion, the DNA replication process could result in errors of interpretation (misdiagnosis) in 7% of proliferating cells. Thus, the use of a cell cycle phase-specific marker could avoid errors by indicating the cell stage in which the nucleus analysed is found. Copyright © 2003 S. Karger AG, Basel.