Much of the work of our archives is done on computer resulting in massive databases which may be inventories of collections, a guide to the collections, bibliographic databases, and the like. I have received numerous requests for copies of the databases on disk so the recipient can search, etc. I have not yet done this....but at first glance, these seemed reasonable. Then I noted some problems with compliance. First, these databases often reflect OUR collections. By allowing them out of our control, this raises the real problem that users may "tinker" with them. Let's say one database is a chronological list of sheet music. A user may find this really useful and start inputting their OWN sheet music into this "catalog." After awhile, the user (and anyone who gets access to the data or information) may come to believe that either 1) the items he/she added are also in OUR collections, or 2) the items from OUR collections are believed to be in the users. What I'm specifically getting here is that the database which lists an institution's collection may mislead later researchers if it is corrupted and passed on. Second, because electronic databases are so easy to manipulate, our lists are constantly changing. The disk I give out in February 1994 may be entirely different than the one in August 1994....unless you really want to keep track of disk "editions," this can become a nightmare, especially if a user begins citing "data" in footnotes, bibliographies, and so on. You also run the continuing risk of misleading users....the old "disk" may lead the person to think that you have no paintings by person X....but you actually do. Third, there is the possibility of corruption of the information. Any narrative information, dates, etc. on the disk bear the possibility of change in the hands of a user. With printed catalogs or guides, such changes can be written or noted and you have therefore "evidence" of change. You may not see that change on the disk. Fourth, your database records may bear information that you really would not give out. For example, donor names, prices paid, dealers, or the fact that you cannot really prove ownership of an item. When you copy electronic data, you essentially copy everything unless you have the time and staff to compile a "public" database as opposed to your "internal" one. Please note that I am truly open to sharing of information. In fact, you can successfully argue that in public institutions where public money has paid the salary of individuals to write/compile these databases...that they are open to public access. That is true, however we do make these databases accessible on site, rather than distribute them. If I could find the means to prevent changes to the electronic information on the disk, then I would openly distribute disk databases. Dean DeBolt, University Librarian Special Collections and West Florida Archives University of West Florida, Pensacola