More fundamental understanding of cell growth regulation will provide novel approaches for detecting, preventing, and treating different cancers. Activation of protooncogenes or loss of tumor-suppressor genes can have both prognostic and therapeutic importance. In epithelial ovarian cancer, poor prognosis is associated with continued expression or overexpression of tyrosine kinase growth factor receptors p170EGFR, p165fms, and p185erbB-2. Over-expression of erbB-2 (HER-2/neu) by breast and ovarian cancers already permits effective targeting of antibodies and immunotoxins. Ultimately, molecular analysis of individual cancers will guide the application of specific therapies to inhibit activated oncogenes or restore the function of tumor-suppressor genes. Circulating growth factors, oncogene products, and tumor-associated antigens may provide markers for earlier detection of some cancers. In epithelial ovarian cancer, concentrations of CA 125 can be increased 1-5 years before clinical diagnosis, and approximately 50% of patients with stage I disease have had an abnormal CA 125 concentration. Combining CA 125 with two novel markers--OVX1 and M-CSF--has retrospectively detected > 98% of stage I ovarian cancers. Although the specificity of the three markers is insufficient for cost-effective screening, serum tests for them could prompt the performance of transvaginal sonography, substantially decreasing the number of sonograms required. Genetic markers in the germ line of patients at increased risk for certain cancers will almost certainly influence strategies for prevention or detection. In familial breast, and ovarian cancer, a locus on chromosome 17q tracks risk of cancer in a fraction of kindreds. How often germline abnormalities will be detected in patients with apparently sporadic cancer remains to be determined.