Abstract: Nuclear transfer was carried out with mammary epithelial cells, fetal fibroblasts and cumulusgranulosa cells by method of intracytoplast injection and electrofusion respectively. The influence of cell types, cell refrigeration and the methods of reconstruction of donor cells to developmental ability of cloned embryos were compared. The donor cells were treated with/without refrigeration, and were synchronized at G0/G1 stage by serum deprivation before nuclear transfer. Results: (1)The developmental ability of cloned embryos originated from mammary epithelial cells (blastocyst rate: 7.14%) was significantly lower than those from fetal fibroblasts (16.19%) and cumulusgranulosa cells(19.01%) (P<0.05). (2) There was no significant difference among three kinds of cell (P>0.05) mentioned above in the influence of cell refrigeration on developmental ability of cloned embryos. (3) The rate of reconstruction with electrofusion（69.52%）(P<0.01）was significantly higher than that with intracytoplast injection (59.20%) to mammary epithelial cells. However, there was no significant difference between these two methods (P>0.05) for fetal fibroblasts. For cumulusgranulosa cells, the rate with intracytoplast injection (82.17%) was significantly higher than that with electrofusion(50.99%) (P<0.05), and the rate of forming blastula was a little higher than that with electrofusion, but not significantly different. Conclusions: Fetal fibroblasts and cumulusgranulosa cells were superior to sustain the developmental ability of cloned embryos. Cell refrigeration had no significant effect on the reconstruction of cloned embryos. Comparing the ways of reconstructing embryos, it was suggested that electrofusion was better for mammary epithelial cells，but intracytoplasm injection was better than electrofusion for cumulusgranulosa cells. Both methods were fit to fetal fibroblasts.