One of the most promising approaches to management of ovarian cancer is early detection. Stage I ovarian cancer can be cured with currently available therapy in more than 90% of patients. However, fewer than 25% of ovarian cancers are currently detected in stage I. Detection of a greater fraction of cancers at an early stage might improve clinical outcome. Given a prevalence of one patient with ovarian cancer among 2,500 asymptomatic postmenopausal women in the general population, a successful screening strategy must have a sensitivity of more than 75% and a specificity of more than 99.6% to achieve a positive predictive value of 10%. Approaches to screening include transvaginal sonography, serum markers, and two-stage strategies that use alterations in serum markers to prompt sonographic examination. Among the serum markers, CA-125 has been studied most extensively. Isolated values of CA-125 lack adequate sensitivity or specificity, but when monitored over time, serial CA-125 values can achieve a specificity of 99.6%. However, sensitivity is limited and CA-125 may only be expressed by 80% of early-stage cancers. Multiple markers may exhibit greater specificity when studied over time. To combine multiple markers, more sophisticated mathematical analysis will be required. At present, screening women at conventional risk should be restricted to clinical trials. In the future, however, screening for ovarian cancer may reduce the morbidity and mortality of this disease.