PURPOSE: The aim of the study was to compare the prognostic value of a response by the Gynecologic Cancer Intergroup (GCIG) Cancer Antigen (CA) -125 response criteria and the Response Evaluation Criteria in Solid Tumors (RECIST) on survival in patients with ovarian carcinoma receiving second-line chemotherapy. PATIENTS AND METHODS: From a single-institution registry of 527 consecutive patients with primary ovarian carcinoma, 131 records satisfied the inclusion criteria: ovarian carcinoma of International Federation of Gynecology and Obstetrics stage IC to IV, first-line chemotherapy with paclitaxel and a platinum compound, refractory or recurrent disease, and second-line chemotherapy consisting of topotecan or paclitaxel plus carboplatin. Univariate and multivariate analyses of survival were performed using the landmark method. RESULTS: In patients with measurable disease by RECIST and with assessable disease by the CA-125 criteria (n = 68), the CA-125 criteria were 2.6 times better than the RECIST at disclosing survival. In a multivariate Cox analysis with inclusion of nine potential prognostic parameters, CA-125 response (responders v nonresponders; hazard ratio, 0.21; P < .001) and number of relapse sites (solitary v multiple; hazard ratio, 0.47; P = .020) were identified as contributory prognostic factors for survival, whereas the parameters of RECIST (responders v nonresponders), as well as the remaining variables, had nonsignificant prognostic impact. CONCLUSION: The GCIG CA-125 response criteria are a better prognostic tool than RECIST in second-line treatment with topotecan or paclitaxel plus carboplatin in patients with ovarian carcinoma.