To determine if chemotherapy dose intensity influences treatment outcome in advanced ovarian cancer, all randomised studies of first line chemotherapy, published between 1975 and 1989, were analysed for relationships between planned dose intensity and (a) objective response and (b) median survival. Total dose intensity of each study regimen was calculated and a weighted regression model providing for systemic differences in response or survival among studies was utilised. Hence, treatment arms of different studies were never directly compared. In addition, relative dose intensities of individual drugs within combinations was similarly evaluated. The improvement in objective response rate when adding one unit of total dose intensity ranged between 12% and 16% depending on baseline response rate. The improvement in median survival when adding one unit of total dose intensity ranged between 2 and 4 months. One unit of total dose intensity corresponds to, for example, 20 mg m2 week of cisplatin, or 25 mg m2 week of doxorubicin, or 350 mg m2 week of cyclophosphamide. The analysis of individual drugs suggested that doxorubicin and the platinum compounds were about equally effective, with cyclophosphamide being less effective. The methodological benefits and limitations of the approach used and the implication of the results are discussed.