The objective of this study was to determine the effect of exogenous mitochondria obtained from granulosa cells on the development of bovine embryos in vitro. We classified cumulus oocyte complexes (COCs) as good (G)- and poor (P)-quality oocytes based on cytoplasmic appearance and cumulus characteristics, and assessed mtDNA copy numbers in the G and P oocytes with real-time polymerase chain reaction (PCR). The mitochondria were isolated by fractionation and suspended in mitochondria injection buffer (MIB). Part one of the experiment consisted of the following treatments: (1) G-oocytes + sperm, (2) P-oocytes + mitochondria + MIB + sperm, (3) P-oocytes + MIB + sperm, and (4) P-oocytes + sperm. In part 2, oocytes were parthenogenetically activated. The treatments were: (1) G-oocytes, (2) P-oocytes + mitochondria + MIB, (3) P-oocytes + MIB, and (4) P-oocytes alone. The results indicated a significant difference in mtDNA copy number between G (361 113 +/- 147 114) and P (198 293 +/- 174 178) oocytes (p < 0.01). The rates of morula, blastocyst, and hatched blastocysts derived from P-oocytes + mitochondria were similar to those of G-oocytes, but significantly higher than P-oocytes without exogenous mitochondria in both the ICSI and parthenogenetic activation experiments. We found no difference in blastomere numbers between G-oocytes and P-oocytes + mitochondria in either experiment, but blastomere numbers in these two groups were significantly higher than in P-oocyte groups without exogenous mitochondria. These data suggest that mtDNA content is very important for early embryo development. Furthermore, the transfer of mitochondria from the same breed may improve embryo quality during preimplantation development.