Establishment of cryopreserved follicle banks would provide a valuable source of immature oocytes for subsequent follicular culture and embryo pro- duction. In order to develop an optimal cryopreservation protocol, a comparison of ovarian and follicular cyropreservation procedures was performed. Whole ovaries or isolated follicles of B6CBAF1 mice were either frozen slowly or vitrified, and post-thaw follicle growth and oocyte maturation, parthenogenetic activation and embryo development after IVF, and follicle ultrastructure were subsequently monitored. In ovary cryopreservation, intrafollicular oocytes (n = 67-80) can mature (n = 20-21; 25-31%) and develop into blastocysts (n = 8-12; 12-16%) after parthenogenesis. When comparing optimal methods, slow freezing using 1.5 M of DMSO with a 0.5-mL straw resulted in more cleavage than did vitrification using 7.5% ethylene glycol (EG) + DMSO with an EM grid (P < 0.05) after IVF. In follicle cryopreservation, slow freezing (n = 79) yielded only mature oocytes (n = 19; 24%), but none were parthenogenetically activated. Major cryoinjury after ovary cryopreservation was intracytoplasmic vacuole formation and mitochondrial deformity. In follicular cells, vitrification induced more damage in mitochondria than did slow freezing, whereas both induced mitochondrial damages and vacuole formation in ooplasm. Slow freezing did not damage zona pellucida. In conclusion, banking of the whole ovary yields better outcome of follicle culture than follicle banking and, compared with vitrification, slow freezing efficiently supports post-thaw survival and ultrastructure normality in ovary cryopreservation. However, each protocol induces cell-specific damage.