A pharmacokinetically guided phase I study of topotecan and etoposide phosphate was conducted in recurrent ovarian cancer. The scheduling of the topoisomerase I and II inhibitors was determined using in vitro activity data. All patients had recurrent disease following prior platinum-containing chemotherapy. Patients had a World Health Organisation performance status of 0–2 and adequate bone marrow, renal and hepatic function. Treatment was with topotecan intravenously for 5 days followed immediately by a 5-day intravenous infusion of etoposide phosphate (EP), with pharmacokinetically guided dose adjustment. Plasma etoposide levels were measured on days 2 and 4 of the infusion. A total of 21 patients entered the study. In all, 48% were platinum resistant and 71% had received prior paclitaxel. The main toxicities were haematological, short lived and reversible. A total of 29% of patients experienced grade 4 thrombocytopenia and 66% grade 4 neutropenia after the first cycle. Neutropenia and thrombocytopenia was dose limiting. The maximum-tolerated dose was topotecan 0.85 mg m−2 day−1 days 1–5 followed immediately by a 5-day infusion of EP at a plasma concentration of 1 μg ml−1. The response rate (RR) was 28% in 18 evaluable patients. There was marked interpatient variability in topoisomerase IIα levels measured from peripheral lymphocytes, with no observed increase following topotecan. This regimen of topotecan followed by EP demonstrated good activity in recurrent ovarian cancer and was noncrossresistant with paclitaxel. Both the toxicity and RR was higher than would be expected from the single agent data, in keeping with synergy of action.