Purpose of investigation. Significant progress has been made in recent years in the understanding of the mechanisms postulated by the gonadotropin theory of ovarian carcinogenesis. In the present study we compare FSH concentrations between serum and fluid from cysts or the rectouterine pouch of patients with epithelial tumors and non-neoplastic lesions. Methods. We enrolled 277 patients. They were divided into five groups: I (n = 44) - ovarian cancer patients, II (n = 16) - borderline tumors, III (n = 40) - benign epithelial cystadenomas, IV (n = 137) - non-neoplastic lesions and V (n = 22) - admitted for “second-look” laparoscopy. Results. There were any significant differences between FSH concentrations in serum and tumor fluid in patients with ovarian cancer (36.46 vs 28.11 mIU/mL) and borderline epithelial tumors (31.5 vs 22.7 mIU/mL). For benign cystadenomas the respective concentrations were 28.96 mIU/mL in serum and 6.93 mIU/ml in tumor fluid in these groups p < 0,0000001. The same highly significant differences were found in non-neoplastic lesions (24.97 vs 4.77 mIU/mL), p < 0,0000001. Patients who underwent “second-look” laparoscopy demonstrated significant differences (p < 0.05) as FSH concentration in serum and peritoneal fluid when neoplastic cells were not disclosed, but the difference was not significant (p = 0.752) when fluid from the rectouterine pouch was positive for carcinoma cells. Conclusions. The results of our study can reflect an ineffective tumor: blood barrier and easy diffusion of gonadotropins into the tumor tissue. Local reduction of FSH levels through administration of GnRH analogs may in some clinical situations produce clear therapeutic benefits for the management of ovarian malignancies.