This study was designed to compare single-point prevalence estimates with a cumulative prevalence estimate of human papillomavirus (HPV) type 16 in cervical smears. The influence of the menstrual cycle and the long-term effect of HPV 16 positivity on the development of cervical intraepithelial neoplasia (CIN) were monitored. We examined 21 women (median age 23.6 years) every 5 weeks for 1 year. All women had a history of negative Papanicolaou smears for at least 5 years before enrollment. Cervical swabs were collected at each visit for Papanicolaou smears and HPV 16 detection by the polymerase chain reaction. Five years after completion of the last visit, self-reported information regarding cervical neoplasia was obtained. Human papillomavirus type 16 point-prevalence estimates per visit varied between 14.3-33.3%. The cumulative prevalence was 66.7%; 14 women were positive at least once and seven women were continuously negative for HPV 16. Detection of HPV 16 was significantly higher in the luteal phase. Repeated positivity for HPV 16 by Southern blot and polymerase chain reaction preceded and accompanied CIN in one patient, whereas in the remaining patients, positivity for HPV 16 by polymerase chain reaction alone was not associated with CIN during a 5-year follow-up. Single-point measurements of HPV 16 in cervical smears by polymerase chain reaction are of limited value for assessment of an individual's HPV status. This should be kept in mind when HPV testing for screening programs is considered.