The relation between reproductive factors and cervical neoplasia was evaluated in a case-control study of 528 cases of invasive cancer compared with 456 control subjects in hospital for acute conditions unrelated to any of the established or suspected risk factors for cervical cancer, and of 335 cases of cervical intraepithelial neoplasia compared with 262 outpatient controls. The risk of invasive cervical cancer increased with number of livebirths, the estimated multivariate relative risk (RR) being 4.39 in women with five or more births compared with nulliparous women. There was also an inverse relation with age at first livebirth (RR = 0.42 for greater than or equal to 30 vs. less than 20 years) which, however, disappeared after inclusion of parity in multiple logistic regression analysis. Likewise, cases of invasive cervical cancer tended more frequently to report induced abortions. However, this association was not statistically significant after allowance for confounding factors, including parity. No relation emerged with number of spontaneous abortion and age at last pregnancy. When the interaction between parity and sexual habits was analysed, the relative risk increased in subsequent strata of parity with increasing number of sexual partners or decreasing age at first intercourse, thus suggesting an independent effect of sexual and reproductive factors, and hence multiplicative on the relative risk of invasive cervical cancer. No consistent association emerged between the risk of intraepithelial cervical neoplasm and parity, number of abortions and age at first or last birth.