Beginning in 1999, a web-based metadata and data repository we call the "open research system" (ORS) was designed and built to assist geographically distributed scientific research teams. The purpose of this innovation was to promote the open sharing of data within and across organizational lines and across geographic distances. As the use of the system continued, end users and group administrators requested the development of a second, Intranet-based system with similar functionality. After three years of operation, a survey was conducted of users of the system to understand why some users and research groups appeared to utilize the two systems more than others. From this research we found that some barriers to use include: (1) mismatch of system functionality to user or group needs; (2) willingness to share with an internal group by Intranet but not with the world by Internet; and (3) resistance to entering metadata because of workplace habits. This experience has also taught us that with respect to web-based metadata and data repositories there is a difference between long-term and short-term research projects in their need to establish good metadata and data storage procedures.