Some previous attempts to produce tetraploids experimentally have resulted in a proportion of treated embryos becoming 2n/4n mosaics at a frequency which may be as high as 20%, when using cytochalasin B as a fusigenic stimulus and cytogenetic techniques to identify putative tetraploid embryos. To investigate the possible occurrence of 4n/2n mosaicism, tetraploid embryos were produced by electrofusion, a process which allows adjacent blastomeres at the 2-cell stage to fuse following exposure to electric field pulses. Embryos used for electrofusion were hemizygous for a transgene consisting of approximately 1000 copies of the mouse β-globin gene. After in situ hybridization, one hybridization signal is expected per diploid genome. Tetraploid cells in 7·5-, 8·5-, 9·5- and 10·5-day-old conceptuses were distinguished from diploid cells by performing in situ hybridization on histological sections. The frequency of nuclei with two hybridization signals in the ‘hemizygous’ tetraploid embryos was compared to diploid embryos which were either hemizygous or homozygous for the β-globin transgene. Comparison of the frequency of nuclei with two hybridization signals between tissues of ‘hemizygous’ tetraploid conceptuses and homozygous diploid conceptuses showed no significant difference, which implies that the tissues in the tetraploid conceptuses were uniformly tetraploid. No evidence was found to suggest that electrofusion results in 2n/4n mosaicism.