Remarkable improvements in primary surgery and chemotherapy for advanced ovarian cancer have been achieved in the last decades. Nevertheless, the majority of patients still develop recurrent disease and ultimately die from ovarian cancer. Evaluation of efficient second-line treatment is of clinical relevance. This review re-analyses the published data of conventional systemic treatment for recurrent and refractory ovarian cancer. Patients were grouped according to prior chemotherapy (with or without platinum; with or without paclitaxel) and according to clinical platinum resistance. Platinum resistance is defined as having progressed during platinum based chemotherapy or having relapsed within 6 months after completion of primary platinum containing treatment. The most favourable results in platinum sensitive ovarian cancer were reported for platinum containing re-induction chemotherapy. Results in platinum refractory ovarian cancer were different. Both hormonal treatment consisting of either tamoxifen or GnRH analogues, and chemotherapy with topoisomerase blocking agents, taxanes, or alkylating agents did show similar results. Unfortunately, there are only limited data from prospectively randomised trials in these patients. Therefore, recommendations remain inconclusive. Retrospective comparisons may help the clinician to chose the currently best available treatment for an individual patient, however, these treatments have to be evaluated in prospectively randomised trials. The protocols of the ongoing studies in refractory or recurrent ovarian cancer of the AGO Ovarian Cancer Study Group are outlined.