Several studies have suggested that the quality of coital data from diaries is superior to that collected by retrospective questionnaires. By collecting data over short intervals of time, diaries can present a more comprehensive picture of exposure, while minimizing the potential for recall bias. Despite these advantages, paper diaries have limited use because of their expense and difficulty of implementation. Web-based data collection offers the opportunity to make improvements to the quality of epidemiologic exposure measurement by providing privacy and convenience to study participants while reducing costs associated with questionnaire administration and allowing for real-time data processing. We adapted coital diaries for Web-based data collection in a study of transmission rates of genital human papillomavirus infection among young adults. University women complete an online sexual behavior questionnaire ("diary") every 2 weeks over a 3-year follow-up period; men complete a single online sexual behavior questionnaire ("journal"). In this paper we describe the design, methodology and implementation issues that emerge in conducting a Web-based epidemiologic study. We also discuss compliance, as well as methods for assuring appropriate security, confidentiality and privacy.