Objective: To establish the evaluation system of the quality of oocytes on the basis of the incidence of cumulus cells apoptosis, to investigate the relationships beween the incidence of cumulus cells and the outcomes of IVF-ET. Method: Thirth-four cycles undergoing controlled ovarian hyperstimulation for IVF-ET with tubal infertility (23 cycles) or unexplained infertility (11 cycles) were included in this study. Cumulus cell masses surrounding mature oocyte and co-culture of embryos with autologous cumulus cells during IVF-ET process. The incidence of apoptosis in cumulus cells was assessed by apoptosis detection kit fluorescein. The effect of co-culture using cumulus cells and the incidence of cumulus cells apoptosis. Results: The results were as follows: 1. The incidence of apoptosis in cumulus cells markedly increased in patients aged 40 or over, while the fertilization rate was greatly decreased in those age group. 2. Apoptosis in cumulus cells was found in both the fertilized oocytes and unfertilized oocytes, but the incidence of apoptosis was higher in unfertilized oocytes. 3. There is no clear correlation between apoptosis in cumulus cells and the number of oocytes retrieved. However, the incidence of apoptosis was increased when the number of oocytes retrieved was 5 and fewer in comparison with $6{sim}10$. 4. Embryo grade was significantly affected by the incidence of apoptosis in cumulus cells. 5. Pregnancy rate of IVF-ET per cycle was 29.4%, and the pregnant group had the higher fertilization rate and a significantly lower incidence of apoptosis in cumulus cells compared with the nonpregnant group. 6. When cumulus cells were used as helper cells in the co-culture of the embryo, in vitro activity of cumulus cells based on morphological change and proliferation did not influence the quality of embryo, but was closely associated with the implantation rate and pregnancy rate, which was enhanced when morphological changes and proliferation of cumulus cells was more active. 7. This difference in the outcome of IVF-ET according to in vitro activity of cumulus cells used for co-cultue was not associated with the incidence of apoptosis in cumulus cells; but rather had likely relations with the different secretion pattern of protein, which may be an embryo trophic factor by cumulus cells. Conclusion: These results suggest that the incidence of apoptosis in cumulus cells can be used in predicting oocyte qualities and the outcomes of IVF-ET. And the effect of co-culture largely depends on the in vitro activity of cumulus cells as well.