As no unobjectionable study has yet been published on ovarian cancer screening with ultrasound and/or tumor markers in the general population or in risk groups, it is not possible to determine if screening would decrease the mortality in ovarian cancer. Because the medical gain is unclear, the cost per life gained can not be estimated. At present, there is not sufficient scientific evidence to recommend screening for ovarian cancer with any method. It may be worthwhile to perform a large randomized trial to discover if early detection is of value, as ultrasound is already being used on a large scale in asymptomatic women in order to detect ovarian cancer. Such a study will also gain knowledge about the early stages of ovarian cancer. Moreover, methods for investigation of lesions detected by ultrasound (i.e., color Doppler and puncture) could be evaluated. Only through a large-population study would we be able to know if prophylaxis for ovarian cancer (oral contraceptives) is more efficient than screening. Studies on lung cancer have shown that persuading people to stop smoking more effectively decreases the mortality of lung cancer than screening. A population study would be a giant enterprise, but probably the only way to determine whether screening for ovarian cancer with ultrasound is of any value. The following attitude to ultrasound examination in risk groups is proposed as long as the value of screening is unclear: Women with hereditary ovarian cancer (one or two first degree relatives with ovarian cancer) should be made aware of the risks, and regular ultrasound check-ups seems advisable for psychologic reasons and because the risk for cancer is high. The surveillance proposed by Lynch seems reasonable: ultrasound at least once a year, possibly in combination with CA-125, and discussion concerning surgical removal of the ovaries after childbearing is complete or around 40 years of age. The protective effect of oral contraceptives should be emphasized. Women at risk should be made aware of the fact that mesothelial cancer can develop even after ovarian removal. If possible, these women should be registered and their family history noted. Thus, their risk of contracting ovarian cancer can be estimated and compared to the outcome when the ovaries are removed. Mesothelial cancers after oophorectomy must be noted.(ABSTRACT TRUNCATED AT 250 WORDS)