During the past few years, several Harvard paper surveys were converted to Web surveys. These were high-profile surveys endorsed by the Provost and the Dean of the College, and covered major portions of the university population (all undergraduates, all graduate students, tenured and non-tenured faculty). When planning for these surveys started in 2001, Web surveys were still avant-garde. Dillman's 2000 edition of his classic "Mail and Internet Surveys: The Tailored Design Method" devoted only 29 of 450 pages to Web surveys, indicative of the small size of the collected body of knowledge about Web surveys at that time. The authors did not have a good practical manual to guide them in the Web survey process, so they wrote their own. They have delivered versions of this paper at various conferences and at each venue, received a big group of professionals seeking ways to get started, or needing tips to refine their own processes. Since 2001, their group has done several Web surveys together at Harvard, using a variety of software packages. They also have drawn upon the growing body of literature now available on Web surveys. In this paper, they will begin farther back in the process with how the survey projects were conceptualized and enacted in a complex university environment. They then take readers through many organizational and technical aspects of doing a Web-survey. Along the way, they talk about some emerging issues surrounding survey research, such as privacy and incentives. (Contains 15 tables and 4 endnotes.)