All Caucasian women in a large Eastern city who developed pathologically confirmed cervical cancer between 1950 and 1969 are being prospectively followed in an epidemiological test of the venereal hypothesis of cervical carcinogenesis. We are attempting to identify all men who were married to these probands at any time prior to the date of their cancer diagnosis. The ultimate objective is the identification of all the other wives of the proband husbands in order that their risk of cervical cancer be assessed. A random sample of control wives similar to the other wives in age, race, date and place of marriage as well as prior marital status is also being followed. To date, a total of 1,087 other wives and 659 control wives has been fully traced. Cervical cancer or carcinoma in situ was detected in 29 (2.7%) of the other wives and in seven (1.1%) of the control wives. A total of 14.0% of the other wives had either cervical cancer or a cervical cytological specimen which was other than normal. The corresponding statistic for the control wives was 8.0%. These differences in the prevalence of cervical cancer and of non-normal cervical cytology are statistically significant. In the course of this investigation so far, we have identified 29 "marital clusters" of cervical cancer in which two women married to the same man have all developed cervical neoplasms. The observed number of 29 clusters may be compared with an expected number of 11.6. This investigation, as yet incomplete, offers confirmatory evidence of the possible role of venereal factors in the pathogenesis of human cervical neoplasia. While the genital herpesvirus is the likeliest candidate, other venereal elements might also be involved.