There have been many reports to date regarding the role of GnRH as a local regulatory factor of ovarian function as studies of human and rat ovaries revealed GnRH and its receptor. In recent studies it has been shown that GnRH directly causes apoptosis in the granulosa cells of the rat ovary, and such results leads to the suggestion that the use of GnRH agonist for more stable long term ovarian hyperstimulation in human IVF-ET programs causes granulosa cell apoptosis which may lead to follicular atresia. Therefore this study attempts to determine if granulosa-luteal cell apoptosis occurs in patients during IVF-ET programs in which GnRH agonist is employed for ovarian hyperstimulation. The quality of oocyte-cumulus complexes obtained during ovum pickup procedures were assessed morphologically and then the fertilization rate and developmental rate was determined. Apoptotic cells among the granulosa-luteal cells obtained during the same procedure were observed after staining with Hematoxylin-rosin. The fragmentation degree of DNA extracted from granulosa-luteal cells was determined and comparatively analyzed. There was no difference in the average age of the patients, the number of oocytes retrieved, and fertilization and developmental rates between the FSH/hMG group and GnRH-long group. There was also no difference in the apoptosis rate and pyknosis rate in the granulosa-luteal cells between the two groups. However, when the oocyte-cumulus complexes were morphoogically divided into the healthy group and atretic group without regard for the method of hyperstimulation, the results showed that the number of oocytes obtained averaged 11.09+/-8.75 and 10.33+/-4.53 per cycle, respectively, showing no significant difference, but the fertilization rate (77.05%, 56.99%, respectively, p<0.01) and developmental ,ate (65.96%, 41.51%, respectively, p<0.01) was significantly increased in the healthy group when compared to the atretic group. The degree of apoptosis in the granulosa-luteal cells showed that in the healthy group it was 2.25% which was not significantly different from the atretic group (2.77%), but the pyknosis rate in the atretic group (27.81%) was significantly higher compared to the healthy group (11.35%, p<0.01). The quantity of DNA fragmentation in the FSH/hMG group was 32.22%, while in the GnRH-long group it was 34.27%, showing no significant difference. On the other hand the degree of DNA fragmentation was 39.05% and 11.83% in the healthy group and atretic group, respectively, showing significantly higher increase in the atretic group (p<0.01). The above results suggest that death of granulosa-luteal cells according to the state of the oocyte-cumulus complex is more related to pyknosis rather than apoptosis. Also, the GnRH agonist used in ovarian hyperstimulation does not seem to directly affect the apoptosis of retrieved oocytes and granulosa-luteal cells, and which is thought to be due to the suppression of the apoptogenic effect of GnRH agonist as a result of the high doses of FSH administered.