Background Cyclo-oxygenase-2 (COX-2), the key enzyme in the conversion of arachidonic acid to prostaglandins, is involved in critical steps of tumor onset and progression, and is a strong predictor of chemotherapy resistance and poor outcome in advanced ovarian cancer. To our knowledge, no data has been reported until now about the association between COX-2 status and response to different chemotherapy regimens. Methods A retrospective study was performed to investigate the association of COX-2 with outcome and response to platinum versus platinum/paclitaxel in 68 primary ovarian cancer. COX-2 immunoreaction was performed on paraffin-embedded sections by using rabbit polyclonal antiserum against COX-2. Results In the overall series, COX-2 positivity was found in a statistically significant higher percentage of not responding cases than in patients responding to chemotherapy (n = 15/21; 71.4% versus n = 17/47; 36.1%; p value = 0.0072). A higher percentage of COX-2 positivity was found in patients unresponsive (n = 11/13; 84.6%) versus patients responsive to platinum-based chemotherapy (n = 9/26; 34.6%). In cases administered platinum/paclitaxel, COX-2 positivity was found in 4 out of 8 (50%) of un responsive versus 8 out of 21 (38.1%) of responsive cases. Logistic regression analysis of parameters likely to affect response to treatment resulted in a p value = 0.17 for the interaction COX-2/type of treatment. Conclusion Although these findings need to be confirmed in a larger series, our study suggests a possible indication that there is a difference in the influence of COX-2 on response depending on treatment regimen.