In this paper we discuss the design implications involved in collecting and processing data via three disparate survey modes: mail, Web, and computer-assisted telephone interviewing (CATI) for the National Survey of Recent College Graduates (NSRCG). We provide a brief background of the study and outline the design and development process. Also discussed are the similarities and differences between the modes, examples of questions that posed particular challenges, and survey management implications. We conclude with lessons learned and advice for future multimode surveys of this nature.