Despite increasing success rate of IVF, poor response to ovarian stimulation remains a problem. So, attempts to improve ovarian responses, for example, by using combined gonadotropin-releasing hormone analogue(GnRH-a) and human menopausal gonadotropin(hMG) have shown limited success. It is reported that response of granulosa cells in vitro to FSH is stimulated by co-incubation with IGF-l, and IGF-l production can be increased by growth hormone. This suggest that combination regimen of G.H. and hMG may augment follicle recruitment. In fifteen patients who had previous history of poor ovarian response to gonadotropin stimulation after pituitary suppression with mid -luteal GnRH-a, the effectiveness of cotreatment with G.H. in IVF program was evaluated using a combination regimen of G.R. and hMG at Korea University Hospital IVF Clinic. Ovarian responses to gonadotropin stimulation in control and GH-treated cycles assessed by total dose and duration of hMG treatment, follicular development and peak $E_2$ level, number of eggs retrieved, and fertilization rates were also assessed. In each group, serum and follicular fluid IGF-1 concentrations on day of egg collection were measured by RIA after acidification and extraction by reveresed phase chromatography. Patients receiving G.H. required fewer days and ampules of gonadotropins, developed more oocytes, and more embryos transferred. But, the differences were not statistically significant, except the duration of hMG treatment. Our data showed a significantly higher concentration of IGF-l in the serum, not in the follicular fluid, of patients treated with G.H. compared with control group. These data suggest that growth hormone treatment does not improve the ovarian response in women with limited ovarian reserve to gonadotropin stimulation for IVF.