The role of reproductive factors in the aetiology of epithelial ovarian cancer has been re-assessed in a meta-analysis of 3 hospital-based case-control studies conducted in Europe (i.e. Italy, the United Kingdom and Greece), providing a total dataset of 1,140 cases and 2,724 controls. Multiple logistic regression models were used to obtain relative risk (RR) estimates adjusted for study centre, age, socio-cultural indicators, age at menopause, and oral contraceptive use. The risk decreased with increasing number of births and the trend in risk was significant (chi 2(1) = 7.50, p less than 0.01). In comparison to nulliparous women, those who reported 4 or more births had a 40% reduction in risk of ovarian cancer (RR = 0.6, 95% confidence interval, CI: 0.4-0.8). An RR estimate of 1.4 (95% CI: 1.1-1.7) as found, overall, for age of 35 or more at first birth compared to age of 25 or less at first birth. In each stratum and overall, nulliparous women did not appear to be at increased risk compared to those who delayed birth of their first child until age 35 or more. In each study, as well as in the overall dataset, an inverse association between number of abortions and ovarian cancer risk emerged. Overall, the inverse relationship was highly significant, RR estimates for 1 and 2 or more abortions, as compared to none, being 0.9 (95% CI: 0.8 and 1.1) and 0.7 (95% CI: 0.6-0.9) respectively. The effects of parity, age at first birth and number of abortions emerged consistently in various strata of study centre and age.