The histoculture drug sensitivity assay (HDRA) has been demonstrated to have high predictability for resistance, sensitivity, and survival for gastrointestinal cancer (Clin Cancer Res 1: 305-311, 1995; Clin Cancer Res 1: 1537-1543, 1995). In this report, we evaluated the clinical usefulness of the HDRA in ovarian cancer. HDRA was performed on tumors from patients with ovarian cancer. Eighty-five cases (97%) were evaluable. Tumor fragments were cultured on collagen-sponge gels. The cultures were incubated with cisplatin (CDDP) for seven days. Cell viability were assessed with the MTT end point. The optimal cut off concentration of CDDP was determined to be 25 micrograms/ml by correlation with the historical clinical response rate to CDDP. HDRA results were correlated to clinical response of 15 patients who received CDDP-based therapy that included doxorubicin and cyclophosphamide (CAP therapy). The true positive rate was 88%, the true negative rate was 86%, the sensitivity was 88%, the specificity was 86%, and the accurate prediction rate was 87% when HDRA results were compared to the response of the treated patients. The data suggest that the HDRA is capable of predicting the response to antitumor chemotherapy in patients with ovarian cancer and that measuring response to CDDP can be useful for optimization of CAP chemotherapy for patients with this disease.