Ovarian cancer is the most frequent cause of death from gynaecological malignancies in the Western world. Most cases of epithelial ovarian cancer are detected at late stages and the resultant overall five-year survival is poor. However, when epithe­lial ovarian cancer is detected with the disease confined to the ovary the prognosis is favorable. Transvaginal gray-scale ultra­sonography and colour Doppler assessment of blood flow have been evaluated as methods to predict risk of malignancy in ovarian tumours. In order to reduce the number of unnecessary surgical procedures for uterine adnexal tumours, ultra­sonomorphologic scoring systems have been developed, assigning numerical ultrasonographic parameters of the tumours However, the positive predictive value of these scoring systems is low and this is due to the fact that the appearance of many benign ovarian lesions overlaps with that of malignant disease. In addition, some ovarian malignancies are ultrasonographically detected as simple cysts without exhibiting a complex morphology. Moreover, the cut-off size of uterine adnexal tumours for surgical intervention in the early detection of cancer is not yet well determined. The application of colour blood­flow imaging is very helpful in the detection of uterine adnexal malignancy because of the presence of neovascularization in malignant tumours. When gray-scale ultrasonography detects the presence of septum or papillary projections or solid com­ponents in uterine adnexal lesions and Doppler flow is present within these lesions malignancy is likely. However, the detec­tion of vascularity within the papillary projection of a malignant tumour may not be detected when it is very small. When colour-flow imaging is used in premenopausal patients attention is needed to avoid confusion of luteal flow with flow of cystic lesions. Initial reports using pulsed Doppler ultrasonography showed high sensitivity and specificity in the detection of ovarian cancer when levels of the resistive index (RI) less than 0.4 and levels of the pulsatility index (Pl) less than 1 were used. Subsequent studies have shown considerable overlap of RI and Pl rates between benign and malignant uterine adnexal masses, suggesting that pulsed Doppler ultrasonography is not an independent indicator for malignancy. Serum CA-125 levels have been used in conjunction with ultrasonography to identify as many of the false-positive results in order to avoid unnec­essary surgery. In postmenopausal women with a uterine adnexal mass the combination of physical examination with serum CA-125 levels and pelvic ultrasound scan seems to improve the sensitivity and specificity of predicting adnexal malignan­cies. In contrast, in premenopausal women the consideration of CA-125 levels with Doppler ultrasonographic findings might confuse the differential diagnosis of ovarian masses. In conclusion, accurate selection of patients with uterine adnexal tumours for surgical intervention is not provided by pelvic ultrasonography. Pelvic ultrasonography as a screening method for the early detetection of ovarian cancer should be probably limited to those women who are at increased risk for devel­opment of ovarian cancer and not in the general population.