The Papanicolaou (Pap)-smear history of 212 cases of invasive cervical cancer was compared with that of 1060 age-matched controls drawn from neighbours. In the 5 years before the year of diagnosis 32% of the cases had been screened by Pap smear, compared with 56% of the controls. This difference was statistically highly significant (p less than 0.0001) and indicated a relative risk of invasive cancer of 2.7 in women who had not been screened by Pap smear, compared with those who had. Differences in Pap-smear history between cases and controls persisted when the data were stratified by age, income, education, marital history, smoking habit, employment status, and access to medical care. These results support the belief that the Pap smear is an effective screening procedure for invasive cervical cancer.