Our objective was to examine whether past infection with human papillomavirus (HPV)-16, as determined by an antibody assay, is a risk factor for subsequent cervical cancer. Incident cases of in situ or invasive cervical cancer occurring between 1975 and 1990 in a cohort of over 11,000 healthy women in Washington County, MD, were identified. The baseline sera of cases and of matched controls, collected in 1974, were examined for IgG antibodies reactive with virus-like particles of HPV-16, a cancer-associated HPV, and HPV-6, a low-risk HPV. Postdiagnosis sera of 11 cases were also assessed similarly. Fourteen cases of invasive and 28 cases of in situ cervical cancer and 83 matched controls were evaluated. The main outcome measure was the risk of cervical cancer in women who had HPV-16 or HPV-6 antibodies in prediagnostic sera. Antibodies to HPV-16 but not to HPV-6 were a marker for subsequent occurrence of cervical cancer. Case sera were reactive more often and more strongly with HPV-16 virus-like particles than were sera of matched controls. The presence of antibodies to HPV-16 was significantly associated with an increased risk of cervical cancer (odds ratio, 3.9; 95% confidence limits, 1.4, 10.7); high antibody levels to HPV-16 were associated with an even greater risk of cervical cancer (odds ratio = 7.5, 95% confidence limits 1.5, 36.3). The association with cervical cancer was strengthened after adjustment for smoking and years of education. In tests of 11 pairs of pre- and postdiagnostic sera, HPV-16 antibodies did not decline markedly over a 7-13-year time period, and seroconversion to HPV-16 appeared to have occurred in 2 cases. The serological data indicate that HPV-16 infection is associated with future risk of cervical cancer and strengthen the evidence for the etiological role of HPVs in cervical cancer.